Search is not available for this dataset
article
stringlengths 4.36k
149k
| summary
stringlengths 32
3.35k
| section_headings
listlengths 1
91
| keywords
listlengths 0
141
| year
stringclasses 13
values | title
stringlengths 20
281
|
|---|---|---|---|---|---|
The long-held principle that functionally important proteins evolve slowly has recently been challenged by studies in mice and yeast showing that the severity of a protein knockout only weakly predicts that protein’s rate of evolution . However , the relevance of these studies to evolutionary changes within proteins is unknown , because amino acid substitutions , unlike knockouts , often only slightly perturb protein activity . To quantify the phenotypic effect of small biochemical perturbations , we developed an approach to use computational systems biology models to measure the influence of individual reaction rate constants on network dynamics . We show that this dynamical influence is predictive of protein domain evolutionary rate within networks in vertebrates and yeast , even after controlling for expression level and breadth , network topology , and knockout effect . Thus , our results not only demonstrate the importance of protein domain function in determining evolutionary rate , but also the power of systems biology modeling to uncover unanticipated evolutionary forces .
Over evolutionary time , every protein accumulates amino acid changes at its own characteristic rate , which Zuckerkandl and Pauling likened to the ticking of a molecular clock [1] . Remarkably , this evolutionary rate varies by orders of magnitude among proteins . Understanding the determinants of this variation is a fundamental goal in molecular evolution research [2–5] . Early theoretical work suggested that functional constraints within proteins [1] and the functional importance of each protein to the organism [6 , 7] would be key factors in determining evolutionary rates . Yet , empirical studies using knockouts have observed only weak effects . In bacteria [8 , 9] , yeast [10 , 11] , and mammals [12] knockout studies conclude that essential proteins evolve only slightly more slowly than non-essential proteins . Moreover , among non-essential genes in yeast , there is little to no correlation between the effect of a protein knockout on growth rate , in a wide range of conditions , and that protein’s evolutionary rate [11 , 13 , 14] , particularly when controlling for expression level [15] . This poor correlation between knockout effects and rates of protein evolution has led some researchers to conclude that function-specific selection plays little role in determining evolutionary rates [4 , 5] . This conclusion , however , contradicts theoretical expectations , the intuition of most molecular biologists , and the reasoning behind much of comparative genomics [16] , motivating our search for an alternative measure of protein function . We reasoned that knockouts do not mimic evolutionarily relevant mutations , which often have small or moderate effects [17] . In particular , most amino-acid changes do not completely destroy a protein’s function , but rather alter its biochemical activity to a greater or lesser extent [18] . The ideal experiment would thus measure the functional effects of many random mutations on many proteins , but such experiments remain challenging [19] . To overcome this experimental limitation , we undertook a computational approach , using biochemically-detailed systems biology models to predict the effects that small perturbations to protein activities will have on the dynamics of the networks in which they function ( Fig 1 ) . We ascribed high and low dynamical influence to protein domains for which amino acid substitutions were predicted to have respectively large or small effects on network dynamics . We hypothesized that network dynamics is a synthetic phenotype that is likely subject to natural selection . To test this hypothesis , we compared our predictions of dynamical influence in functionally and structurally conserved intracellular signaling and biosynthetic networks with genomic data on protein domain evolutionary rates in both vertebrates and yeast . We found that , within these networks , dynamical influence is as strongly correlated with evolutionary rate as many previously known correlates . Moreover , dynamical influence remains predictive when knockout phenotype , expression , and network topology are controlled for . Dynamical influence thus offers new insight into selective constraint within dynamical protein networks .
Biochemically-detailed systems biology models encapsulate vast amounts of molecular biology knowledge in a form that can be used for in silico experimentation [20 , 21] . In particular , they enable simulation of the dynamics of molecular species ( e . g . , proteins , metabolites , modified forms , and complexes ) concentrations under a variety of conditions . In these models , protein biochemical activities are quantified by reaction rate constants k [22] . To assess the phenotypic effects of small changes in protein activity caused by mutations , we first calculated the dynamical influence of each reaction rate constant ( Eq 1 , Materials and Methods ) . To do so , we calculated how a differential perturbation to that constant would change the concentration time course of each molecular species in the network ( Fig 1D ) , for biologically-relevant stimuli . We then normalized those changes and integrated the squared changes over time . Lastly , we summed over all molecular species in the network . The dynamical influence of a rate constant is thus the total effect that small changes in that rate constant would have on network dynamics . The dynamical influence of each reaction rate constant quantifies its importance to network dynamics , but there is little data on evolutionary divergence of reaction rate constants to which we can compare . To compare with the abundant genomic data detailing sequence divergence at the domain level , we aggregated the influences of reaction rate constants for all reactions in which a given protein domain is involved . Whenever possible , we analyzed at the domain level , because that is the level at which distinct functions can be assigned to distinct regions of protein sequence [23] . Thus , we defined the dynamical influence D of a domain to be the geometric mean of the dynamical influences of the reaction rate constants for reactions in which it participates ( Fig 1A , Eq 2 ) . In general , any mutation in a domain will cause a multidimensional perturbation of all rate constants associated with that domain . Furthermore , different mutations in a domain will differ in the overall magnitude of that perturbation and its relative effect on different parameters [24] . Unfortunately , little systematic data exists about the distributions of such perturbations . The geometric average we took is an approximation to the more complex averaging that occurs as various mutations arise over evolutionary time . As more systematic data is generated about mutation effects on biochemical activities of different domains [19] , the geometric average may be replaced by domain-specific distributions of perturbations . To test whether dynamical influence is informative about protein evolution , we analyzed dynamic protein network models from BioModels [25] , a database which not only collects such models but also annotates them with links to other bioinformatic databases [26 , 27] . We considered only models with experimental validation that were formulated in terms of molecular species and reactions , were runnable as ordinary differential equations , and contained at least eight distinct UniProt protein annotations . In total , we studied 12 vertebrate [28–39] and 6 yeast [40–45] signaling and biosynthesis models . We further annotated these models to connect molecular species and reactions with particular protein domains ( S1 Dataset ) . For each model , we calculated dynamical influences for each reaction rate constant using the stimulation conditions considered in the model’s original publication ( S1 Text ) . Using this novel method , we determined protein domain dynamical influence and evolutionary rate for 18 conserved signaling and metabolic networks ( Fig 2 ) . We quantified the strength of the relationship between dynamical influence and evolutionary rate using Spearman rank correlations ( ρ ) , and in 10 of 12 vertebrate networks and 6 of 6 yeast networks , we found a negative correlation . This is consistent with the expectation that most sequences and networks evolve primarily under purifying selection [46] , in which natural selection is primarily acting to remove deleterious mutations from the population . Mutations in protein domains with high dynamical influence are predicted to have greater phenotypic effect and thus , in general , be more deleterious . So mutations in those domains are more efficiently removed , and those domains evolve more slowly . Demonstrating the strength of our approach , the two exceptional vertebrate models with a positive correlation , visual signal transduction and interleukin 6 ( IL-6 ) signaling , were recently identified as undergoing network-level adaptation in humans using population genetic data [47] . Positively selected molecular changes in rhodopsin associated with changes in absorption wavelength have been shown to affect dose-response behavior in visual signal transduction [48 , 49] , suggesting that network-level adaptation may compensate for changes in rhodopsin . As part of the innate immune system , IL-6 and its receptor evolve under strong diversifying selection , so downstream proteins may evolve to maintain signal fidelity . Moreover , viruses are known that directly interfere with proteins downstream of IL-6 [50 , 51] , potentially driving additional adaptation . Dynamical influence is thus predictive not only about purifying selection but also about adaptive selection . The strength of the correlation between dynamical influence and protein domain evolutionary rate varies considerably among networks ( Fig 3A , S1 and S2 Tables ) . Dynamical influence quantifies the relative effects of perturbations within a single network . Directly comparing dynamical influences among networks would require assumptions about the heterogenous relative fitness impact of those networks , such as EGF/NGF signaling versus cell cycle progression . Pooling of heterogeneous data can lead to biased estimates of overall correlations [53 , 54] . To avoid such bias , we did not pool data across networks but rather applied meta-analysis . Because selection may act differently on networks with different functions , we considered a random-effects meta-analysis . Thus the sampled networks were assumed to represent a population of networks , among which the correlation between domain evolutionary rate and dynamical influence varies . The meta-analysis seeks to estimate the distribution of those correlations . We applied the random-effects meta-analysis method of Hunter and Schmidt [52] , because simulation studies suggest that it provides an accurate estimate of the mean correlation , particularly when that correlation is modest [55 , 56] . The estimated distribution of correlations between domain evolutionary rate and dynamical influence is wide , but the 95% confidence interval for the mean correlation excludes zero ( Fig 3A and Table 1 ) . This suggests that negative correlation is more common than positive correlation , consistent with the expectation that purifying selection is more common than adaptive selection [46] . As a complementary approach to evaluating the relationship between domain evolutionary rate and dynamical influence , we also performed a permutation test for dependence between them . In this test , we compared the estimated mean correlation from the real data with a null distribution of mean correlations calculated from scrambled data ( Materials and Methods ) . This test rejected the null hypothesis that domain evolutionary rate and dynamical influence are independent within networks , consistent with the confidence interval analysis that excludes zero correlation ( Table 1 ) . We measured dynamical influence using hand-built systems biology models; what effect do uncertainties in these models have on our analysis ? To be agnostic about what aspects of network dynamics are critical to fitness , in calculating dynamical influence we summed over the integrated sensitivities of all molecular species in the network . It is , however , often evident that the builders of each model had specific molecular species in which they were most interested . If we restricted our dynamical influence calculation to those species ( S1 Text ) , we found very similar correlation with domain evolutionary rate ( Fig 4A ) . Our results are thus not strongly sensitive to which aspects of network function are assumed to be subject to natural selection . Given a network model , substantial uncertainty can exist about the values of the rate constants k [22] , because they are difficult to measure directly and are thus often fit to experimental data on network behavior [57 , 58] . To account for this rate constant uncertainty , an ensemble of rate constant sets consistent with the experimental data and the model can be built [59 , 60] , but this has unfortunately been done for only a small number of models . To assess the importance of rate constant uncertainty to our results , we used an ensemble of 2000 sets of rate constants [61] that were previously identified as consistent with experimental data for one of our models of EGF/NGF signaling [28] . This ensemble was built by Markov Chain Monte Carlo sampling of the posterior distribution when fitting the model to data from 14 systems biology experiments . Rate constant values in the resulting ensemble vary dramatically , with many values varying by more than four orders of magnitude , but all sets of rate constants reproduce the experimentally-measured network dynamics . We calculated the dynamical influence of all protein domains in the network using all these sets of rate constants . Comparing 10 , 000 randomly chosen pairs of sets of dynamical influences to each other , we found that they were highly correlated ( Fig 4B ) , with a median rank correlation of 0 . 74 . Over the ensemble of plausible rate constant sets , the correlation between domain dynamical influence and evolutionary rate varied in magnitude , but 99 . 8% of rate constant sets yielded a negative correlation ( Fig 4C ) . Together these analyses suggest that , while rate constants themselves vary dramatically over the ensemble for this model , relative dynamical influence varies much less , such that rate constant uncertainty does not affect the sign of the observed correlation , although it may affect its magnitude . We could not build rate constant ensembles for the other models in our analysis without access to the original data used to fit those models , but the universality of the “sloppy” pattern of sensitivities in systems biology models [61 , 62] suggests that similar results would be found using rate constant ensembles for the other models in our analysis . In addition to rate constant uncertainty , different modelers may also make different assumptions when studying the same network regarding forms of interaction , which molecular players to include , or which conditions to consider . We assessed the effect of these assumptions using the models in our data which consider overlapping protein domains . The rank correlation between dynamical influences calculated for the same domains using different models varied considerably and was stronger for pairs of models with larger numbers of overlapping domains ( Fig 4D and S3 Table ) . Weighting correlations from different comparisons as in our meta-analysis , we found a mean correlation of 0 . 26 . For comparison , the correlation between different research groups in measurements of gene expression in log-phase growth of budding yeast is roughly 0 . 62 [63] , while for degree in protein-protein interaction data , the correlation is 0 . 11 [64] . Thus model uncertainty plays a strong but not dominant role in our analysis , and it is comparable to variables that have previously been found to be informative about evolutionary rate . We defined dynamical influence in terms of differential perturbations to reaction rate constants ( Eq 1 ) , but mutations introduce finite perturbations . Dynamical influence values calculated using finite perturbations of ±25% to rate constants were , however , almost perfectly correlated with values calculated using differential perturbations ( S4 Table ) . Our results thus also apply to mutations of moderate effect . Dynamical influence captures the phenotypic effect within dynamical networks of small perturbations to protein domain activity , but how does it relate to factors previously linked to evolutionary rate ? In many cases , previously known factors were discovered and validated using genome-wide analyses . The set of protein sequences for which dynamical influence can be calculated is smaller and potentially biased . First , dynamical influence considers effects on network dynamics after stimulus , so it is not applicable to proteins that do not respond to any stimuli . Second , we calculated dynamical influence from mathematical models , and such models exist for only some systems . Lastly , we calculated dynamical influence at the domain level , so we did not consider the evolution of linker sequences between domains . Because the previously known factors we consider are defined at the whole-protein level , they can never fully explain evolutionary rate at the domain level . Nevertheless , we used correlation analysis to understand how dynamical influence compares with previously known predictors of evolutionary rate , for the set of networks and protein domains represented in our study . In multicellular organisms , proteins that are expressed in more cell types ( i . e . , have higher expression breadth ) evolve more slowly [65] , and this is true in the vertebrate networks we study ( Table 1 ) . The significant positive correlation between dynamical influence and expression breadth ( Table 1 ) suggests that protein domains with key roles in these networks exert their effects across multiple tissues , providing a functional explanation for the observed correlation between expression breadth and evolutionary rate . Expanding from expression breadth , the strongest known correlate with protein evolutionary rate is expression level . Proteins with greater expression evolve more slowly in both yeast [66] and vertebrates [67] , which may reflect the costs of protein mis-folding [15 , 68–70] or mis-interaction [71] . In our networks , we found the expected negative correlation between evolutionary rate and expression level ( Table 1 ) . That estimated mean correlation is weaker than that between evolutionary rate and dynamical influence ( Table 1 ) , although the confidence intervals overlap . Indeed , dynamical influence is not significantly correlated with expression level ( Table 1 ) , indicating that dynamical influence reveals previously unanticipated evolutionary pressures beyond the strongest previously known correlate . A significant advantage of our approach is that it captures how molecular inputs are integrated into functional phenotypic outcomes that may be selected upon . One aspect of network biology that has been previously considered in determining protein evolution is topology . Specifically , proteins with more interaction partners ( i . e . , greater degree ) [72] or more central locations within networks ( greater betweenness centrality ) [73] evolve more slowly . Consistent with this previous work , we find that domain evolutionary rate has a significant negative correlation with both protein degree and betweenness centrality ( Table 1 ) . But , intriguingly , dynamical influence of protein domains is not significantly correlated with degree or betweenness centrality ( Table 1 ) of the corresponding proteins . Why is the influence of topology not captured in our dynamics-based analysis of evolutionary rate ? Network topology is a crude measure of function; networks with the same topology can have different dynamics and thus different functions [74] . Thus , our focus on network dynamics rather than topology provides novel insight into protein domain evolution by directly quantifying system output . Expression level and network topology are also represented in the models themselves , by the total abundance of the molecular species that represent each protein and by the reactions that connect them . The correlations of dynamical influence and evolutionary rate with these model-derived quantities ( S5 Table ) are similar to those with experimentally-derived expression and topology ( Table 1 ) . In fact , the partial correlation between dynamical influence and evolutionary rate controlling for expression and topology is stronger when using model-derived values than when using experimental values . The relationship we find between dynamical influence and evolutionary rate is thus not driven by hidden co-variation between dynamical influence and abundance or topology within the models . These assessments of dynamical influence relative to known contributors to protein evolution clearly indicate that our approach has uncovered previously unappreciated constraints on protein evolution . Is this new insight sufficient to explain the conundrums raised by knockout experiments ? In our data , we found that the correlation between evolutionary rate and knockout measures of function was so weak as to be nonsignificant ( Table 1 ) , consistent with prior work [11 , 12] . Strikingly , across the eleven vertebrate networks that include both essential and non-essential proteins and the six yeast networks ( for which knockout growth rate data is available ) , we find no statistical correlation between dynamical influence and essentiality or knockout growth rate ( Table 1 ) . Thus , the highly significant correlation between dynamical influence and evolutionary rate ( Fig 2 , Table 1 ) provides a new perspective on the influence of protein function on evolutionary rate . But , evolutionary rates are complex and likely integrate selection on multiple processes [2–5] . To assess the power of our approach in comparison with alternative integrative analyses , we used partial correlation analysis [12] . Across all our networks , we find that when expression , network topology , and knockout effect are controlled for , the mean correlation between protein domain evolutionary rate and dynamical influence remains statistically significant ( Fig 3 and Table 1 ) . Because the predictive power of dynamical influence cannot be explained by other factors , it provides novel and previously inaccessible insight into evolutionary rates within protein networks . The existence of overlapping protein domains might inflate statistical significance in our analyses across models . To account for this , for all domains that appeared in more than one model , we randomly kept each domain in one of the models and deleted it from the others . We did this randomization one thousand times and repeated our correlation analysis each time , obtaining distributions of mean correlations and permutation p-values . There was a tail of large p-values for the fully controlled correlation between dynamical influence and evolutionary rate , corresponding to randomizations in which many domains happened to be retained in the few models with a positive correlation ( S1 Fig ) . The median results we found were , however , similar to our analysis using all the data ( S6 Table ) , suggesting that overlapping domains do not substantially affect our statistics . Dynamical systems biology models offer great promise for developing and testing evolutionary hypotheses [21 , 75] . Previous topological and flux-balance analysis of networks has offered insight into protein evolution [76–78] , but dynamical models contribute substantial biological detail not previously captured by these approaches . We have shown that incorporating that detail can , for domains within dynamical networks , explain the previous lack of correlation between protein function and evolutionary rate . Dynamical models have previously been used to predict the phenotypic effects of mutations [79] and to assess the correlation between network sensitivity and protein evolution in phototransduction [80] and in pyrimidne biosynthesis [81] . Here we consider many networks to reveal a previously unexplored and general link between dynamical influence and protein domain evolutionary rate within networks . Given the rapid pace of progress in systems biology modeling [82] , the anticipated advances in model scope and validation will provide even more robust data sets to uncover previously unanticipated factors influencing evolutionary processes .
We defined the dynamical influence κi of reaction rate constant ki by κ i 2 = ∑ stimulation conditions c ∑ molecular species y 1 T c ∫ 0 T c d y c ( t ) d k i k i y max 2 k = k * d t . ( 1 ) Here yc ( t ) is the time course of molecular species y in condition c , evaluated using the rate constant values k* from the original publication . The derivative dyc ( t ) /dki of the time course with respect to rate constant ki measures how sensitive that molecular species or metabolite is to changes in that rate constant . To make relative comparisons , we normalized these sensitivities by the value ki of the rate constant and the maximum ymax of molecular species y over all stimulation conditions . We normalized by ymax rather than using a control coefficient d y ( t ) d k k y ( t ) = d ln y ( t ) d ln k [83] because many molecular species in signaling models begin with zero concentration , so the control coefficient would be undefined . We found the total effect of changes in ki by squaring these normalized sensitivities , integrating over the time course of each stimulation condition , and summing over all molecular species and stimulation conditions . We defined the dynamical influence Dd of protein domain d to be the geometric mean of the influences κ of the Nd reaction rate constants for reactions in which that domain participates: D d = ∏ r = 1 N d κ r 1 N d . ( 2 ) We took a geometric mean because rate constant sensitivities range over orders of magnitude [61] . Stochastic noise plays an important role in many cellular networks [84 , 85] . In those cases , networks are not well-modeled by ordinary differential equations ( ODEs ) , although parameter sensitivities can be defined and calculated [86] . To minimize the complications introduced by stochasticity , we focused on models of signaling and biosynthesis in which concentrations of molecular species were sufficiently large to justify a continuous approximation to probabilistic biochemical reaction rates [57] . Moreover , because sources and levels of gene expression noise are known to vary according to initial conditions [87] , we restricted our analysis to models which were fit to experimental data arising from multiple initial conditions and measuring multiple reporters . We downloaded systems biology models in Systems Biology Markup Language ( SBML ) format [88] from the Feb . 8 , 2012 release of BioModels [25] . We calculated dynamical influence for all protein-related biological parameters in each model , using SloppyCell [89] and simulating under the conditions considered in each model’s original paper ( S1 Text ) . These parameters represent a variety of biological phenomena , such as binding and catalytic constants and rates of diffusion and production . We considered only those parameters representing rates of biochemical reactions that depend on protein structure , because we expected constraint on those reactions to have the strongest effect on evolutionary rates . Given the dynamical influences κ for each reaction constant , we reviewed the literature to determine the protein domain or domains at which the reaction occurs , and we assigned those influences to that domain or domains ( S1 Dataset ) . UniProt protein ID’s were acquired from the BioModels annotation in the SBML file for each model and converted to NCBI Protein IDs for vertebrates or open reading frame ( ORF ) numbers for yeast . Some models specified more than one Uniprot ID for a single protein , in cases where there is more than one transcript identified and both appear to perform the same function ( for example , MEK1 and MEK2 ) . Where more than one Uniprot ID was specified , we reviewed the model publication and the protein network literature to select a single transcript . In the case of metabolic flux models that track metabolites rather than proteins , we used the names of the enzymes involved in the reactions to find the appropriate protein identifier . Vertebrate homologous protein alignments were downloaded from the NCBI Homologene database [90] , and for each protein in the alignment , nucleotide sequence was downloaded from NCBI Entrez [90] . These nucleotide sequences were then used as a template to back-translate the Homologene protein alignments to nucleotide alignments . Yeast gene information for the 7 species in the tree in S2 Fig was downloaded from the Saccharomyces Genome Database [91] on Nov . 19 , 2012 . These gene sequences were translated to protein amino acid sequences using Biopython [92] , aligned using ClustalW [93] , and then back-translated to aligned nucleotide sequence using the gene sequence as a template . Protein domain annotation was done manually using literature review , based on information for the human protein in vertebrate models or the Sa . cerevisiae protein in yeast . Evolutionary rates were calculated using codeML from PAML Version 4 . 4b [94] , with one dN/dS ratio per tree ( model 0 ) , the F3x4 codon substitution model , and a rooted tree , as in [95] . The Mgene = 3 setting of codeml was used to estimate a single dN/dS ratio per annotated protein domain . We required a minimum of 4 homologs to include a gene in the analysis , and for each gene any species with more than one homologue was excluded . Because instability is a concern when estimating multiple dN/dS ratios for a single protein sequence , we iterated each codeml run until we acquired three models for which the log-likelihood was within 0 . 01 of the lowest log-likelihood obtained and then used the model with the lowest log-likelihood . Vertebrate gene expression and tissue specificity data was compiled from the mouse GNF1M dataset [96] , downloaded from http://bioGPS . org/downloads . The data consist of microarray probes for a number of tissue types , with each probe’s name including the corresponding gene name , which we mapped to Ensembl gene IDs using Ensembl BioMart [97] . We restricted our analysis to normal adult tissues as in Fig S2 of [95] . To calculate the expression level corresponding to each microarray probe , we took the arithmetic average over replicates of the same tissue and then took the geometric average over tissues . To calculate the expression level of each gene , we then took the arithmetic average of the probe expression levels corresponding to that gene . Yeast expression data [98] was obtained from http://younglab . wi . mit . edu/pub/data/orf_transcriptome . txt and used without modification . Protein abundance within each model was calculated as the sum of initial conditions for all molecular species corresponding to a given protein , including modified forms and complexes . None of the models we considered included transcription or translation , so total levels of all proteins were constant throughout the simulations . We downloaded mouse knockout phenotype data from the Mouse Genome Informatics database [99] at http://www . informatics . jax . org/phenotypes . shtml on July 11 , 2011 . We assembled phenotype information for homozygous knockouts and coded a gene as essential if it resulted in one of the following phenotypes: abnormal reproductive system physiology , prenatal lethality , perinatal lethality , postnatal lethality , premature death , abnormal reproductive system morphology , lethality at weaning , preweaning lethality , partial lethality , and all sub-phenotypes of these phenotypes . If homozygous knockout of a gene did not cause one or more of these phenotypes we coded it as non-essential . To validate our parsing of this data , we compared against the results of [12] . Data for yeast knockout growth rate on YPD media were obtained from the file Regression_Tc1_hom . txt downloaded from the Stanford YDPM database http://www-deletion . stanford . edu/YDPM/YDPM_index . html on March 13 , 2013 . We downloaded protein-protein interaction data for both humans and yeast from the Interologous Interaction Database [100] on April 20 , 2012 . These data take the form of a list of interactions between two proteins , and the dataset from which the interaction was curated . Because we were interested in experimentally verified interactions we restricted our analysis to the HPRD , BIND , IntAct , and INNATEDB datasets for humans and the Krogan_Core , Yu_GoldStd , YeastHigh , YeastLow , and BIND datasets for yeast . We used the python package NetworkX [101] to load these lists of interactions and compute each protein’s degree and its betweenness centrality , which is the fraction of all of the shortest paths between protein pairs in the network that pass through that protein . Model-derived network degree and centrality were calculated from a graph with nodes for each domain in the model and edges between any pairs of domains that participate in a reaction together . Dynamical influence and evolutionary rate are defined at the domain level , but all other factors in Table 1 are defined at the protein level , so in our statistical analyses these other factors were assumed to be equal for all domains within a given protein . We used partial correlation analysis to assess the degree to which these factors account for the observed relationship between dynamical influence and domain evolutionary rate . To do so , within each model we fit linear models for dynamical influence and evolutionary rate as a function of all the other factors and then computed the product-moment correlation between the residuals from these models . Only domains for which all variables were measured were included in these partial correlations . To analyze correlations and partial correlations across models ( Table 1 , Fig 3 , and S5 Table ) , we applied the random-effects meta-analysis approach of Hunter and Schmidt [52] . In this approach , the mean correlation ρ0 in the population is estimated by the average of the observed correlations r of the sampled models , weighted by the sample size n of domains with relevant data in each model . So the estimated mean correlation is ρ ^ 0 = ∑ n i r i ∑ n i , ( 3 ) where sums here and below are over models ( Eq . 3 . 1 in [52] ) . The variance across samples from the population σ r 2 is the sum of the variance in population correlations σ ρ 2 and the variance due to sampling error σ e 2: σ r 2 = σ ρ 2 + σ e 2 . ( 4 ) The variance across samples can be estimated as [52] σ ^ r 2 = ∑ n i ( r i - ρ ^ 0 ) 2 ∑ n i . ( 5 ) The sampling variance can be estimated as ( Eq . 3 . 5 in [52] ) σ ^ e 2 = ∑ n i ( 1 - r i 2 ) 2 n i - 1 ∑ n i . ( 6 ) The standard deviation σρ of the population correlations sets the width of the distribution curves in Fig 3 , and it can be estimated by solving Eq 4 for σρ and substituting in the estimates σ ^ e and σ ^ r . The standard error S E ( ρ ^ 0 ) of the estimated mean correlation ρ ^ 0 depends on the number of population samples K , which is 18 here ( Eq . 5 . 1 in [52] ) : S E ( ρ ^ 0 ) = σ ^ r / K . ( 7 ) The 95% confidence intervals reported in Table 1 and S5 Table are thus ρ ^ 0 ± 1 . 96 S E ( ρ ^ 0 ) . The most popular alternative approach for random-effects meta-analysis of correlations , developed by Hedges and colleagues [102] , estimates the population mean correlation ρ0 and the standard error of that estimate S E ( ρ ^ 0 ) using more complex weightings based on Fisher’s r-to-Z transform . We adopted the Hunter and Schmidt approach because simulation studies suggest that it produces more accurate estimates of the population mean correlation and more accurate confidence intervals when variation in the population is large [56] . In our permutation tests , our null model was that dynamical influence or evolutionary rate was uncorrelated with other protein domain properties ( Table 1 ) . To generate null distributions of correlations , we permuted dynamical influences , evolutionary rates , and all other factors within each model . Because domains share reactions , their influences are not independent , and thus we could not simply permute them to simulate our null model . Instead , we permuted the influences of reaction parameters , which are the most basic unit of our analysis , and we then recalculated the influence for each domain based on the new sets of parameter influences . Evolutionary rates are defined at the domain level , so we simply permuted them within each model . The other factors are defined at the protein level , and we permuted them at the protein level , so that domains within the same protein would still always , for example , have the same expression level . To generate the null distribution for the partial correlation between domain evolutionary rate and dynamical influence , controlling for other variables , we permuted the residuals from the linear models used to calculate the partial correlation [103 , 104] . This approach disrupts any relationship between domain evolutionary rate and dynamical influence while preserving all other relationships between network variables . We permuted variables or residuals separately within each model and then used Eq 3 to calculate mean correlations across models for each permutation . After carrying out 10 , 000 permutations , the two-sided p-values we report ( Table 1 ) are the quantiles of the real data absolute mean correlations among the permuted absolute mean correlations . In all cases , the permutation test results are compatible with the 95% confidence intervals ( Table 1 ) ; smaller p-values correspond to confidence intervals that more strongly exclude zero . Note that permutation tests based on correlation or partial correlation coefficients are strictly tests of the null hypothesis that the two variables are independent [105] . Thus our permutation test cannot reject the possibility that the considered variables are dependent but uncorrelated .
|
Different proteins evolve at dramatically different rates . To understand this variation , it is necessary to determine which characteristics of proteins are visible to natural selection and how the strength of selection depends on those characteristics . One protein characteristic that is evidently visible to natural selection is expression level; more highly expressed proteins are subject to stronger purifying selection and evolve more slowly . Theory and intuition suggest another such characteristic should be some measure of functional importance , but studies of various measures of functional importance , such as knockout essentiality or knockout growth rate , have shown at best weak correlations with evolutionary rate . Here we develop a novel measure of functional importance , dynamical influence , which quantifies the importance of a protein or protein domain to the dynamics of the network of proteins in which it functions . Using 18 biochemically-detailed systems biology models , we compute dynamical influences for each protein domain in each model . We find that dynamical influence is indeed visible to natural selection and that within networks protein domains with higher dynamical influence evolve more slowly .
|
[
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"vertebrates",
"animals",
"systems",
"science",
"mathematics",
"statistics",
"(mathematics)",
"fungal",
"evolution",
"discrete",
"mathematics",
"combinatorics",
"research",
"and",
"analysis",
"methods",
"mycology",
"evolutionary",
"rate",
"computer",
"and",
"information",
"sciences",
"proteins",
"mathematical",
"and",
"statistical",
"techniques",
"statistical",
"methods",
"molecular",
"evolution",
"systems",
"biology",
"biochemistry",
"permutation",
"meta-analysis",
"protein",
"domains",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"evolutionary",
"biology",
"evolutionary",
"processes",
"organisms"
] |
2016
|
Selection on Network Dynamics Drives Differential Rates of Protein Domain Evolution
|
Protozoan parasites belonging to genera Leishmania and Trypanosoma are the etiological agents of severe neglected tropical diseases ( NTDs ) that cause enormous social and economic impact in many countries of tropical and sub-tropical areas of the world . In our screening program for new drug leads from natural sources , we found that the crude extract of the endophytic fungus Cochliobolus sp . ( UFMGCB-555 ) could kill 90% of the amastigote-like forms of Leishmania amazonensis and inhibit by 100% Ellman's reagent reduction in the trypanothione reductase ( TryR ) assay , when tested at 20 µg mL−1 . UFMGCB-555 was isolated from the plant Piptadenia adiantoides J . F . Macbr ( Fabaceae ) and identified based on the sequence of the internally transcribed spacer ( ITS ) regions of its ribosomal DNA . The chromatographic fractionation of the extract was guided by the TryR assay and resulted in the isolation of cochlioquinone A and isocochlioquinone A . Both compounds were active in the assay with L . amazonensis , disclosing EC50 values ( effective concentrations required to kill 50% of the parasite ) of 1 . 7 µM ( 95% confidence interval = 1 . 6 to 1 . 9 µM ) and 4 . 1 µM ( 95% confidence interval = 3 . 6 to 4 . 7 µM ) , respectively . These compounds were not active against three human cancer cell lines ( MCF-7 , TK-10 , and UACC-62 ) , indicating some degree of selectivity towards the parasites . These results suggest that cochlioquinones are attractive lead compounds that deserve further investigation aiming at developing new drugs to treat leishmaniasis . The findings also reinforce the role of endophytic fungi as an important source of compounds with potential to enter the pipeline for drug development against NTDs .
Protozoan parasites belonging to the genera Leishmania and Trypanosoma ( order Kinetoplastida , family Trypanosomatidae ) occurs in the tropical and sub-tropical areas of the world , where they cause severe diseases with huge medical , social , and economic impact to millions of people [1] . All diseases caused by these parasites are among the Neglected Tropical Diseases ( NTDs ) listed by the World Health Organization [1] . Different species of Leishmania affects over 12 million people and puts over 350 million people at risk in 88 countries; Trypanosoma cruzi infects approximately 8 million and puts 100 million at risk in Central and South America , and T . brucei infects 60 million people in 36 sub-Saharan African countries [2] . The drugs currently available to treat the different forms of leishmaniasis and trypanosomiasis were introduced many decades ago and have significant drawbacks , especially in terms of efficacy , length of treatment , route of administration , toxicity , and cost [2] . To complicate the situation , there is no new drug being developed by the major pharmaceutical industries for these diseases [3] . It is well known that plant-associated microorganisms produce a variety of metabolites with novel structures and interesting biological activities [4] , [5] , [6] , [7] . Indeed , some medicinal properties and biological activities initially attributed to plant species were found latter to be due to the secondary metabolites produced by their endophytic microorganisms [8] . With the aim to discover new drug leads for some NTDs , we have been bioprospecting the Brazilian flora and mycota using bioassays which includes the protozoan parasite Leishmania amazonensis and the enzyme trypanothione reductase ( TryR ) . This flavoenzyme is part of a complex enzymatic system present in protozoans of the order Kinetoplastida that help them to survive under oxidative stress [9] . More important , TryR was shown to be essential for the growth and survival of these parasites , and was validated as a drug target for the discovery and design of new leishmanicidal and trypanocidal drugs [9] , [10] , [11] , [12] . In our bioprospecting program , we have prepared hundreds of extracts of endophytic fungi isolated from plants growing in different Brazilian biomes ( unpublished results ) . The isolate UFMGCB-555 , obtained from Piptadenia adiantoides J . F . Macbr ( Fabaceae ) , showed strong activity in the assay with TryR and L . amazonensis and was chosen for investigation aiming at its bioactive components . In this report we describe the molecular taxonomic identification of UFMGCB-555 and the isolation and identification of two leishmanicidal compounds from its extract .
The endophytic fungus was isolated from Piptadenia adiantoides J . F . Macbr ( Fabaceae ) and deposited at Coleção de Microrganismos e Células da Universidade Federal de Minas Gerais under the code UFMGCB-555 . The fungus was grown in potato dextrose agar ( PDA ) medium for 5 days at 28±2°C . Five millimeter diameter plugs of this culture were then inoculated at the center of 160 Petri dishes ( 90 mm diameter ) , each containing 20 mL of malt extract agar ( MEA ) medium ( 1% glucose , 1% malt extract , 0 . 1% peptone , and 1 . 5% agar ) . The plates were incubated at 28±2°C for 14 days . After this period , a small aliquot of the biomass was used for extraction of DNA and the remaining material extracted with ethyl acetate for the isolation of the fungal secondary metabolites . Five plates which were not inoculated with the fungus were subjected to the same protocol to serve as control of the culture medium . The DNA was extracted according to the protocol described by de Hoog [13] . The ribosomal DNA internal transcribed spacer domains ( rDNA-ITS ) were amplified using the primers ITS1 ( 5′-TCCGTAGG-TGAACCTGCGG-3′ ) and ITS4 ( 5′-TCCTCCGCTTATTGATATGC-3′ ) , as previously described [14] . Five sequences were generated using MEGABACE ( Amersham Biosciences , USA ) which were used to feed PHRED-PHRAP software [15] in order to find the consensus sequence . The sequence thus obtained was compared with those deposited in the GenBank using the software BLASTn [16] to identify the fungus to the genus level . The sequence was deposited in the GenBank ( accession number EU684269 ) . Phylogenetic relationships were calculated using the version 4 . 0 of the software MEGA [17] . The phylogenetic tree was constructed using the neighbor joining algorithm with bootstrap values calculated from 1000 replicate runs . The Kimura 2-parameter model [18] was used to estimate the evolutionary distances . The fungus culture material from 160 Petri dishes ( approximately 3 liters ) was transferred to a six-liter Erlenmeyer containing 3 . 5 L of ethyl acetate and left in contact for 48 at room temperature . After decantation , the organic phase was filtered and the solvent evaporated under vacuum in a rotary-evaporator at 45°C . Residual solvent was eliminated in a vacuum centrifuge at 40°C , and the crude extract thus obtained was stored at −40°C until use . A similar extraction procedure was carried out using the five plates containing medium only to generate a control extract of the medium . The initial fractionation step involved high-speed counter-current chromatography ( HSCCC ) , using a Pharma-Tech chromatograph model CCC-1000 equipped with three multilayer coils totaling a volume of 850 mL . With the rotor stopped , the coils were filled with the lower aqueous phase of the biphasic mixture of hexane-ethyl acetate-methanol-water ( 1 . 2∶0 . 8∶1∶1 ) . The coils were then rotated at 1000 r . p . m . and the upper phase was pumped at a flow-rate of 5 mL min−1 in tail-to-head direction until hydrodynamic equilibrium was reached , that is , until only the mobile phase was flowing out of the column . Part of the extract ( 800 mg ) was then dissolved in 10 mL of equal parts of the upper and lower phases of the biphasic solvent mixture and this mixture injected into the column . A total of 158 fractions of 15 mL each was collected . They were pooled into 40 groups based on their similarity , as assessed by thin-layer chromatography ( TLC ) on silica gel plates ( Merck ) . Group 10 was fractionated on reversed-phase high performance liquid chromatography ( HPLC ) using a semi preparative column ( 250×20 mm ) filled with RP-18 ( octadecyl ) silica gel with 5 µm average particle size . The separation was run using a gradient of methanol-water from 70 to 100% in 50 min , at a flow rate of 10 mL min−1 . The column effluent was monitored with a UV detector set at a wavelength of 220 nm . Two pure compounds , 1 ( 14 mg ) and 2 ( 2 . 4 mg ) were isolated . Proton ( 1H ) and carbon ( 13C ) nuclear magnetic resonance ( NMR ) spectra ( Figure S1 ) , Distortionless Enhancement by Polarization Transfer ( DEPT ) , Heteronuclear Multiple-Quantum Coherence ( HMQC ) , and Heteronuclear Multiple Bond Correlation ( HMBC ) experiments were run on a Bruker DRX 400 spectrometer using the pulse programs provided by the manufacturer . The substances were dissolved in perdeuterated solvents containing 0 . 1% tetramethylsilane as the internal chemical shift standard . The data obtained from these spectra are summarized in the Tables 2–4 . Mass spectra ( MS ) were acquired on a Thermo Finnigan LCQ-Advantage spectrometer equipped with an electrospray ion ( ESI ) source . Solutions of the compounds at 200 µg mL−1 in MeOH-H2O ( 1∶1 ) were infused at 25 µL min−1 , and the positive and negative mass spectra acquired with a m/z range between 50 and 1000 daltons . The cone voltages were optimized for positive and negative ion analysis in the range between 25 and 50 V . The capillary voltages were set at 4 . 5 kV in positive ion mode and −3 . 1 kV in negative ion mode . In the MS/MS experiments , the parent ion isolation width was 3 . 8 daltons and the normalized collision energy was set at 30% for both compounds 1 . Fifty scans from 150 to 600 daltons were collected to generate the averaged spectra . The TryR microtitre plate assay procedure based on in situ Ellman's-reagent-mediated regeneration of trypanothione , described by Hamilton et al . [19] , was used during the screening and the bioassay-guided fractionation protocols . The assay was performed in 96-well plates ( Costar 9017 , Corning , USA ) using Hepes buffer ( 40 mM , pH 7 . 5 ) with 1 mM EDTA . Each assay well ( 250 µL ) contained enzyme ( 1 mU ) , trypanothione ( 1 µM ) and NADPH ( 200 µM ) . The extracts , fractions and pure compounds were added to the above mixture and incubated at 30°C during 30 min . After this period , Ellman's reagent [5 , 5′-dithiobis ( 2-nitrobenzoic acid ) –DTNB] was added ( 70 µM ) and the absorbance ( Abs ) measured at the wavelength of 412 nm in the kinetic mode during the time ( t ) of 10 minutes at every 10 seconds . The slope of the curve ΔAbs/Δt is proportional to DTNB reduction which , in absence of competing reactions , is proportional to the enzyme activity . The inhibition of the coupled system was calculated as the ratio between slope ΔAbs/Δt ) of the experimental wells and that of the controls without drug , that is , percent inhibition = [1−slopeexp/slopecontr]×100 . The classical assay based on the measurement of NADPH consumption was performed in 1 mL cuvettes ( 1 cm light path length ) at 27°C using a Beckman DU spectrophotometer . Each cuvette contained 500 µL Hepes buffer ( 40 mM , pH 7 . 5 ) , 1 mM EDTA , 4 mU enzyme , 200 µM NADPH , 50 µM cochlioquinone A , and 100 µM trypanothione . The enzyme , NADPH and the compound were pre-incubated for 5 minutes at 27°C . The absorbance measurement at the wavelength of 340 nm was started in the kinetic mode ( 1 reading per second ) for about 30 seconds . The sample compartment cover was opened during 3 to 4 seconds , just enough for the quick addition of the substrate . After closing , the absorbance measurement was continued for another 30 seconds . The initial reaction velocity ( v0 ) was calculated by the instrument software using the first 5–20 data points after the substrate addition . These points should fit a strait line with R2 ( squared correlation coefficient ) greater than 0 . 99 . To estimate the effect of the compound on enzyme activity , the v0 values obtained in the experiments with and without cochlioquinone A were compared . Each experiment was repeated three times . The effect of the extract and isolated compounds on the survival and growth of the human cancer cell lines UACC-62 ( melanoma ) , MCF-7 ( breast ) , and TK-10 ( renal ) , was determined using a colorimetric method developed at the National Cancer Institute-USA [20] , [21] . Briefly , the cells were inoculated in 96-well plates and incubated at 37°C for 24 h in 5% CO2 atmosphere . The solutions of the test samples were added to the culture wells to attain the desired concentrations , and the plates incubated for further 48 h . Trichloroacetic acid was added to each well to precipitate the proteins , which were stained with sulforhodamine B . After washing out the unbound dye , the stained protein was dissolved with 10 mM Tris , and the absorbance measured at the wavelength of 515 nm . Results were calculated using the absorbance measured in the test-wells ( T ) in comparison with that of the control wells corresponding to the initial cell inoculum ( Ti ) and cells grown for 48 h without drug ( Tf ) , using the formula: [ ( T−Ti ) / ( Tf−Ti ) ]×100 . This formula allows the quantification of both growth inhibition ( values between zero and 100 ) and cell death ( values smaller than zero ) . Each sample was tested in duplicate in two independent experiments . Promastigotes of L . amazonensis ( strain IFLA/BR/196/PH-8 ) were obtained from lesions of experimentally infected hamsters . The parasites were incubated for 9 days at 26°C in Schneider's medium buffered at pH 7 . 2 . The promastigotes were then stimulated to differentiate into amastigote-like forms by rising the incubation temperature to 32°C and lowering the pH of the medium to 6 . 0 . After 7 days under these conditions 90% of the parasites were in the amastigote-like forms . The parasite concentration was adjusted to 1×108 cells mL−1 , and 90 µL added to each well of 96-well plates , followed by 10 µL of the solutions containing the samples and control drug ( 0 . 2 µg mL−1 Amphotericin B - Fungisone Bristol-Myers Squibb , Brazil ) . The plates were incubated at 32°C for 72 h and the number of parasites estimated using the MTT ( methyl thiazolyl tetrazolium ) -based colorimetric assay [22] . The results were calculated from the measured absorbancies using the formula [1− ( Absexp/Abscontr ) ]×100 , which express the percentage of parasite death in relation to the controls without drug . All samples were tested in duplicate and the experiments repeated at least once . Experiments to determine the dose response curves and the EC50 ( effective concentration to kill 50% of the parasites ) were run as above , using 1∶2 serial dilutions of the test compounds to reach the appropriate concentrations . The experiments were run in duplicate and repeated at least once . The antimicrobial activity of the samples was evaluated using the following microorganisms: Candida albicans ATCC18804 , C . krusei ATCC2159 , Staphylococcus aureus ATCC12600 , Escherichia coli ATCC25922 and Cryptococcus neoformans ATCC32608 . The yeasts were grown on agar Sabouraud ( Difco ) at 28°C for 24 h , and their inocula were adjusted in saline solution to Mac Farland optical density scale 1 [23] before seeding into plates containing agar Sabouraud . The bacteria were grown in Brain Heart Infusion ( BHI , Difco ) in 6-mL tubes , and their concentration adjusted to 103 to 104 cells mL−1 before inoculation into plates containing agar BHI . In each plate , five clean filter-paper disks with 6 mm diameter were placed equidistant from each other on the surface of the medium . Solutions of the samples at 10 mg mL−1 were prepared in 1% aqueous dimethyl sulfoxide ( 1% aq . DMSO ) . Five microliters of these solutions , corresponding to 50 µg of the sample , were applied to the paper disks , and the plates incubated at 37°C for a period of 24 to 48 h . Aqueous solutions of amphotericin B and chloramphenicol at 10 mg mL-1 were used as positive controls for yeasts and bacteria , respectively . Solvent control consisted of 1% aq . DMSO . The sample was considered active if it caused a growth inhibition halo around the disk to which it was applied . The software GraphPad Prism version 4 . 03 was used to calculate the EC50 values using the non-linear curve fitting of two ore more independent experimental datasets to a four-parameter logistic dose-response model . No constraints were applied to the curve fitting calculations .
A small aliquot ( 1 g ) of the fungus culture was used to isolate the fungal DNA for sequencing of the ITS domains . These sequences were used for the taxonomic identification of the fungus and for the elucidation of its phylogenetic relationships ( Figure 1 ) . The fungus UFMGCB-555 was then identified to the genus level as Cochliobolus sp . and showed a close phylogenetic relationship with C . melinidis ( Genebank access number AF452445 ) , from which its consensus sequence differs by only 2 . 7% ( 12 nucleotides , 70% bootstrap value ) . The culture material ( 3 liters ) yielded 2 g of the crude extract , representing a yield of approximately 0 . 6 g L−1 of culture broth . This extract showed activity in different bioassays ( Table 1 ) when tested at 20 µg mL−1 , but was completely inactive against C . albicans , C . krusei , S . aureus , E . coli and C . neoformans when tested at 50 µg per disk ( data not shown ) . The extract of the control ( sterile ) culture medium was not active in these assays . Approximately 800 mg of the crude extract was subjected to counter-current chromatography in an HSCCC to afford 158 fractions , which were pooled into 40 groups ( Figure 2 ) . After testing all groups in the TryR assay , only Group 10 and Groups 24–40 showed some activity . Groups 24–40 were not studied further because they presented low masses and were constituted of complex and instable mixtures of highly polar compounds . Fraction 10 ( 100 mg ) was further fractionated using reversed-phase HPLC to yield compounds 1 ( 14 mg; 1 . 4% w/w of the crude extract ) and 2 ( 2 . 4 mg; 0 . 24% w/w of the crude extract ) . Both compounds strongly inhibited the growth of amastigote-like forms of L . ( L . ) amazonensis , with EC50 values ( effective concentration to kill 50% of the parasites ) of 1 . 7 µM ( 95% confidence interval = 1 . 5 to 1 . 9 µM ) and 4 . 1 µM ( 95% confidence interval = 3 . 6 to 4 . 7 µM ) , respectively ( Figure 3 ) . However , only 1 was active in the DTNB-coupled TryR assay . Furthermore , while the crude extract was toxic to three human cancer cell lines used in this work ( MCF-7 , TK-10 , and UACC-62 ) , neither Group 10 nor compounds 1 and 2 showed activity against these lineages ( Table 1 ) . The compounds were also inactive against the five pathogenic microorganisms investigated ( data not shown ) . The electrospray ionization mass spectra ( ESI-MS ) of both compounds exhibited quasi-molecular sodiated ion peaks [M+Na]+ with m/z 555 daltons in positive ion mode , and [M-H]− with m/z 531 daltons in negative ion mode , indicating they both have molecular weight of 532 g mol−1 ( Figure 4 ) . The peak integration areas in the 1H NMR spectra ( Table 2 and Table 3 , Figure S1 ) indicated the presence of 44 hydrogen atoms , while the 13C NMR spectra ( Table 4 ) showed 30 signals for each compound . Edited DEPT sub-spectra showed signals due eight methyl groups , one belonging to an acetyl group , and five methylene carbon atoms for both compounds . Signals due to eight methyne carbon atoms , four of them oxygenated , were observed for compound 1 , while seven signals of methyne carbon atoms , three of them oxygenated , were detected for in the spectra of compound 2 . Altogether , these data is compatible with the molecular formula C30H44O8 , with an index of hydrogen deficiency of 18 , corresponding to 9 unsaturations . The analysis of the 1H and 13C spectra allow to infer that both compounds have five double bonds in their structures , with the remaining unsaturations attributed to the presence of four rings . A quinonoid ring in 1 was suggested by the presence of the signals at 181 . 64 and 188 . 86 δ , attributed to carbonyl groups , together with four signals between 134 . 32 and 151 . 40 δ in the 13C NMR spectrum . A phenol moiety in 2 was indicated by the presence of six signals between 107 . 0 and 181 . 5 δ in the 13C NMR spectrum . A signal at 198 . 5 δ suggested the presence of a ketone carbonyl in 2 . The connections between carbon and hydrogen atoms in the structures were established based on the analysis of the two-dimensional NMR experiments ( COSY , HMQC and HMBC - Table 2 and Table 3 ) . The spectral data of 1 and 2 are in agreement with those published in the literature for cochlioquinone A and isocochlioquinone , respectively ( Figure 5 ) .
Several endophytic fungi were isolated from the P . adiantoides , a plant species selected due to the activity of its extract in a panel of assays used to screen the Brazilian flora for bioactive natural products ( unpublished results ) . Among the fungi isolated from this plant , the isolate UFMGCB-555 showed strong activity in the assays with TryR and L . amazonensis . Using molecular taxonomy techniques , we were able to identify this fungus as Cochliobolus sp . ( Pleosporaceae , Ascomycota ) . This genus comprises approximately 50 species occurring all over the world [24] , many of which can parasitize plants and cause considerable agricultural losses [25] . Some species of Bipolaris , the anamorphic state of Cochliobolus , are also the etiologic agent of several human diseases , such as sinusitis , ocular infections , peritonitis , and meningoencephalitis [26] , [27] . However , in the present work we looked for the ability of Cochliobolus UFMGCB-555 to produce secondary metabolites with biological or pharmaceutical potential , especially for neglected tropical diseases . To guide the fractionation process aiming at the isolation of the active compounds of this extract , we decided to use the TryR assay developed by Hamilton et al . [19] . Besides being more economic due to low consumption of the expensive substrate trypanothione , this assay is simpler , safer and faster to perform than the assay with L . amazonensis . Thus , the bioassay with the parasite was used only at the end of the isolation procedure , with the pure compounds . Using this strategy we arrived at a fraction containing two major substances , one active ( 1 ) and the other inactive ( 2 ) in the TryR assay ( Table 1 ) . Both compounds were , however , active against L . amazonensis , showing EC50 values of 1 . 7 µM and 4 . 1 µM , respectively ( Figure 3 ) . After detailed analysis of the ESI-MS and NMR spectra , and comparison of our data with those published in the literature [28] , compounds 1 and 2 were identified as cochlioquinone A and isocochlioquinone , respectively ( Figure 5 ) . However , slightly different interpretation of some NMR signals , as compared with those described by Miyagawa [28] , are noteworthy: a ) the hydrogen at C-5 in both compounds ( Table 2 and Table 3–position 5 ) couples with the three hydrogen atoms bound to C-27 and with the hydrogen at C-4 , thus appearing as a double quartet , with coupling constants J = 7 . 0 and 5 . 1 Hz , and not as quintet , as described by Miyagawa; b ) the two hydrogen atoms at C-2 in compound 1 ( Table 2–position 2 ) show distinct signals centered at 1 . 45 δ and 1 . 08 δ and not one centered at 1 . 35 δ; c ) the same is true for compound 2 ( Table 3–position 2 ) , where the corresponding signals are observed at 1 . 45 and 1 . 08 δ and not only at 1 . 25 δ , and finally; d ) the two hydrogen atoms at C-20 in compound 2 ( Table 3–position 20 ) resonate at 1 . 46 δ and 1 . 74 δ and not only at 1 . 60 δ as described by those authors . Cochlioquinones and related compounds are known to occur in fungi from different genera , including Cochliobolus and Bipolaris [29] , [30] , [31] , Helminthosporium [32] , Drechslera [33] , and Stachybotrys [30] . However , this is the first report on the leishmanicidal activity of compounds belonging to this class of natural products . It is known that quinonoid compounds with free positions in the quinone ring are susceptible to nucleophilic attack by thiol groups to form stable Michael addition products [34] . In this regard , after 1 was unequivocally identified as cochlioquinone A , its activity in the Ellman's-reagent-mediated TryR assay was questioned due to the presence of a quinonoid ring in its structure . Thus , rather than directly inhibiting the enzyme in the assay , cochlioquinone A could be capturing the reduced substrate resulting in the interruption of the chemical reduction of Ellman's reagent ( Figure 6 ) . To confirm this possibility , we performed the assay using the classical protocol based on the measurement of NADPH consumption [35] while using an excess trypanothione ( 100 µM ) in comparison to cochlioquinone A ( 50 µM ) . Indeed , under these conditions no inhibition of the enzyme activity could be observed ( data not shown ) . In view of the above results , we can conclude that cochlioquinone A and isocochlioquinone A are exerting their leishmanicidal effect by hitting targets other than TryR within the parasite . A literature search disclosed the following information: a ) cochlioquinone A is a competitive inhibitor of the ivermectin binding site in Caenorhabdites elegans , with an inhibition constant of 30 µM [36]; b ) ivermectin is also active in vivo against different species of Leishmania [37] , [38]; c ) the mode of action of ivermectin in nematodes is related to its high affinity to glutamate-gated chloride channels , causing an increase in the permeability of the cell membrane to chloride ions [39]; d ) the TDR database of potential drug targets for NDTs ( http://tdrtargets . org ) reveals four genes of L . major expressing putative chloride ions transporters ( LmjF01 . 0180 , LmjF04 . 1000 , LmjF32 . 3370 , and LmjF33 . 1060 ) ; e ) at least three of these genes have orthologs in C . elegans ( C07H4 . 2 , R07B7 . 1 ) , while LmjF01 . 0180 has an ortholog also in T . cruzi ( Tc00 . 1047053504797 . 140 ) . Based on these pieces of information it is plausible to speculate that chloride ions transporters may also serve as a target for cochlioquinone A and related compounds in Leishmania and Trypanosoma . This hypothesis needs further experimental evidences to be confirmed or refuted . Concerning isocochlioquinone A , the literature [40] show that it can inhibit the growth of the malaria-causing protozoan parasite Plasmodium falciparum with IC50 values of 1 . 4 µg mL−1 for the K1 strain ( resistant to chloroquine and pyrimethamine ) , and 3 . 3 µg mL−1 for the NF 54 strain ( susceptible to standard antimalarials ) . These values are close to the EC50 shown against L . amazonensis in the present work . Besides the significant activity against L . amazonensis , our data indicate that 1 and 2 present some degree of selectivity , as they were inactive in the assays with three human cancer cell lines ( Table 1 ) and five pathogenic microorganisms ( data not shown ) used in this investigation . The low toxicity of 1 and 2 to mammalian cell lines reported in this work is in agreement with recently reported data showing that isocochlioquinone A has only a small effect on HeLa and KB cells [29] , [41] . Another study [33] showed that when cochlioquinone A was tested in vitro against different kinases it showed selective activity against diacylglycerol kinase , both in vitro and in whole cell assay employing BW5147 T cell lymphoma lineage . Related compounds , such as cochlioquinone A1 , exhibited selective toxicity towards bovine aortic endothelial cell when compared with normal and cancer cell lines [29] leading the authors to suggest that it may serve for developing new therapeutic agents for angiogenesis-related diseases . The activity in the low micro molecular range towards L . amazonensis and the selectivity of 1 and 2 are reported here for the first time and justify further investigations on compounds of this class to assess their in vitro and in vivo effect on parasites of the genera Leishmania and Trypanosoma . Finally , by disclosing the leishmanicidal activity of two secondary metabolites from an endophytic fungus , the present work reinforces the role of these organisms as an important source of drug lead candidates for the development of new chemotherapeutic agents for NTDs .
|
Protozoans belonging to genera Leishmania and Trypanosoma are single-cell organisms that can infect humans and cause disfiguring lesions and debilitating or fatal diseases , with enormous social and economic impact in many tropical and sub-tropical areas of the world . The drugs currently available to treat the different forms of leishmaniasis and trypanosomiasis were introduced many decades ago and have significant drawbacks , especially in terms of efficacy , length of treatment , route of administration , toxicity , and cost . In our screening program for new natural products with leishmanicidal activity , we found that the crude extract of a fungus living within the plant Piptadenia adiantoides could kill 90% of the amastigote-like forms of Leishmania amazonensis . The bioassay-guided fractionation of the extract resulted in the isolation of cochlioquinone A and isocochlioquinone A , which showed EC50 values ( effective concentrations required to kill 50% of the parasite ) of 1 . 7 µM and 4 . 1 µM , respectively . These compounds were not active against three human cancer cell lines ( MCF-7/mammary , TK-10/renal , and UACC-62/melanoma ) , indicating some degree of selectivity towards the parasites . Our results suggest that cochlioquinones may serve as starting points for developing new drugs to treat leishmaniasis and reinforce the role of endophytic fungi as an important source of natural products with relevant biological activities .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"chemistry/organic",
"chemistry",
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"infectious",
"diseases/protozoal",
"infections",
"biochemistry/small",
"molecule",
"chemistry",
"biochemistry/drug",
"discovery"
] |
2008
|
Leishmanicidal Metabolites from Cochliobolus sp., an Endophytic Fungus Isolated from Piptadenia adiantoides (Fabaceae)
|
Although Leishmania parasites have been shown to modulate their host cell's responses to multiple stimuli , there is limited evidence that parasite molecules are released into infected cells . In this study , we present an implementation of the change mediated antigen technology ( CMAT ) to identify parasite molecules that are preferentially expressed in infected cells . Sera from mice immunized with cell lysates prepared from L . donovani or L . pifanoi-infected macrophages were adsorbed with lysates of axenically grown amastigotes of L . donovani or L . pifanoi , respectively , as well as uninfected macrophages . The sera were then used to screen inducible parasite expression libraries constructed with genomic DNA . Eleven clones from the L . pifanoi and the L . donovani screen were selected to evaluate the characteristics of the molecules identified by this approach . The CMAT screen identified genes whose homologs encode molecules with unknown function as well as genes that had previously been shown to be preferentially expressed in the amastigote form of the parasite . In addition a variant of Tryparedoxin peroxidase that is preferentially expressed within infected cells was identified . Antisera that were then raised to recombinant products of the clones were used to validate that the endogenous molecules are preferentially expressed in infected cells . Evaluation of the distribution of the endogenous molecules in infected cells showed that some of these molecules are secreted into parasitophorous vacuoles ( PVs ) and that they then traffic out of PVs in vesicles with distinct morphologies . This study is a proof of concept study that the CMAT approach can be applied to identify putative Leishmania parasite effectors molecules that are preferentially expressed in infected cells . In addition we provide evidence that Leishmania molecules traffic out of the PV into the host cell cytosol and nucleus .
Leishmaniasis is a disease that affects over 12 million people in approximately 88 countries . It manifests as varying types of cutaneous lesions , mucocutaneous lesions or visceral disease; the type of disease presentation is dependent on both the parasite species and characteristics of the host that are not completely defined . Parasite lesions are sites of inflammation where Leishmania-infected cells are exposed to a range of potential leishmanicidal activities . It is therefore no surprise that Leishmania infected cells exhibit altered responses including refractoriness to IFN-gamma activation [1] , [2] , inhibition of LPS induced signaling [3] , altered calcium mobilization [4] and non responsiveness to inducers of cell death [5] . Together , these responses enable the parasite to persist within the host cell . Studies on other intracellular microorganisms have shown that within the intracellular milieu , intracellular organisms elaborate molecules that target host cell functions . For example , in the apicomplexan organism Toxoplasma gondii , which resides in PVs that are mostly of parasite origin , molecules from secretory organelles have been shown to traffic beyond the pathogen vacuole and target activities in the host cell cytosol and nucleus [6] . Similarly , prokaryotic pathogens such as Salmonella and Legionella synthesize effector molecules in the intracellular milieu that subsequently gain access to the host cell cytosol via type III and type IV secretion apparatus , respectively [7] , [8] . Importantly , these effector molecules are not expressed by organisms that are grown in liquid broth . Presuming that Leishmania species also elaborate molecules that target host cell processes , there is currently limited knowledge of the identity of such molecules and the conditions under which these molecules are synthesized and released into the infected cell . However , there is considerable evidence that Leishmania parasites also differentially express molecules in response to changes in their environment . For example , promastigote stage-specific molecules such as GP46 , GP63 and lipophosphoglycan are rapidly turned off once parasites are internalized into macrophages or when promastigote forms are incubated in specialized media and growth conditions more suited for amastigote growth [9–1] . Also , LIT1 the ZIP family iron transporter was shown to be preferentially induced in parasites that reside within PVs only after several days of infection [12] . Following from these observations , it is plausible to expect that , as compared to axenically cultured organisms , parasites that grow within infected cells or inside infected hosts most likely elaborate new ( or antigenically modified ) molecules that permit them to function in the intracellular environment . It is to be expected though that the differential expression of such molecules would depend , in large part , on the actual details of parasite culture . To identify parasite molecules that are preferentially expressed in infected cells , we elected to implement the change mediated antigen technology ( CMAT ) , which is a variation of the in vivo induced antigen technology ( IVIAT ) . Both of these screens are immunological screens that take advantage of the fact that antibodies can be raised to new molecules which are expressed by organisms that sense changes in their environment when they enter infected hosts as compared to organisms grown under standard laboratory culture [13] , [14] . The immune serum that is elaborated in response to pathogens within infected hosts will include antibodies that are reactive to parasite structural molecules and housekeeping molecules that perform vital metabolic functions in the parasite . Incubation of this immune serum with axenic cultured organisms should result in strong reactivity of the antiserum with many of the latter parasite molecules . In the CMAT and IVIAT screens , immune serum is incubated with lysates of the pathogen grown under standard laboratory conditions to deplete the serum of antibodies that are reactive to parasite structural and housekeeping molecules . The depleted serum becomes enriched for immunoglobulins that are reactive to molecules preferentially expressed within the infected host . To date , CMAT and IVIAT have identified over 1000 in vivo induced antigens in infections including Bacillus anthracis , Tannerella forsythia , Mycobacterium tuberculosis , Salmonella enterica serovar Typhi , group A Streptococcus , E . coli O157:H7 , Vibrio vulnificus and V . cholera , among others [15]–[21] . These newly identified antigens include new virulence factors in these organisms as well as pathogen proteins previously shown to be preferentially expressed in the infected host . In the study presented here , serum reactive to infected cells was raised by immunization of mice with lysates of cells infected with L . pifanoi or L . donovani parasites . After adsorbing this serum extensively against axenic amastigote forms to remove antibodies that are reactive to molecules expressed by axenically cultured organisms , the adsorbed serum was used to screen inducible parasite genomic expression libraries . A few of the endogenous genes identified by the CMAT screen were analyzed for their time-course of expression and their distribution in infected cells . Evidence is presented that Leishmania parasite molecules identified by CMAT traffic outside the PV and are therefore putative candidate effectors .
Leishmania pifanoi promastigotes ( MHOM/VE/57/LL1 ) obtained from the ATCC were grown in Schneiders medium supplemented with 10% fetal calf serum and 10 µg/ml gentamicin at 23°C . L . donovani strain 1S-CL2D from Sudan , World Health Organization ( WHO ) designation: ( MHOM/SD/62/1S-CL2D ) was obtained from Dr . Debrabant ( USDA , MD ) . Promastigotes of this parasite strain were grown in Medium-199 ( with Hank's salts , Gibco Invitrogen Corp . ) supplemented to a final concentration of 2 mM L-glutamine , 100 µM adenosine , 23 µM folic acid , 100 IU and 100 µg/ml each of penicillin G and streptomycin , respectively , 1×BME vitamin mix , 25 mM Hepes , and 10% ( v/v ) heat-inactivated ( 45 min at 56°C ) fetal bovine serum , adjusted with 1 N HCl to pH 6 . 8 at 26°C . Generation of amastigotes forms was carried out as described [22] . L . donovani axenic amastigotes parasites were maintained in RPMI-1640/MES/pH 5 . 5 medium at 37°C in a humidified atmosphere containing 5–7% CO2 in air . L . pifanoi amastigotes were maintained in the amastigote medium above at 34°C . The RAW 264 . 7 murine macrophage cell line was cultured in RPMI supplemented with 10% fetal calf serum and 100 units Penicillin/Streptomycin at 37°C under a 5% CO2 atmosphere . Genomic DNA of L . donovani and L . pifanoi parasites was prepared using the TELT method [23] . The genomic DNA was digested with Sau3A and 0 . 5–2 . 0 kb fragments were size selected . BL21 ( DE3 ) -based inducible expression libraries were constructed with L . pifanoi and L . donovani fragments in pET30abc using techniques previously described [13] , [24] . Macrophages infected for 24 H with amastigotes of L . donovani or L . pifanoi were lysed in lysis buffer ( 25 mM HEPES , 150 mM NaCl , 1% Triton X-100 , 1x protease inhibitor pill ( 1 pill/10 mL ) , 20 ul/mL Phosphatase Inhibitor Cocktail 1 , 20 ul/mL Phosphatase Inhibitor Cocktail 2 ( Sigma ) , 50 uM NaVO4 ) . 100 ug of lysate per mouse was emulsified in an equal volume of titermax gold adjuvant ( Sigma Aldrich ) . BALB/c mice were immunized and boosted twice . Sera from 4 mice was combined and used in adsorptions . Antibodies reactive with antigens expressed by axenically grown parasites were removed by 4 successive incubations of sera with whole parasites ( 5×108 parasites per adsorption ) in PBS supplemented with 1% bovine serum albumin ( BSA ) and 0 . 05% sodium azide as described earlier [24] . Sera were further incubated with parasites lysed by French Press ( 4×109 parasites ) and heat treated parasites that had been charged onto microspheres ( polystyrene microspheres , Bangs Labs , Fishers IN ) . An additional adsorption was performed with RAW 264 . 7 lysates charged onto microspheres . Aliquots were saved after each adsorption Reactivity of adsorbed sera was determined by incubation of serially diluted sera in 96 well plates ( Nalge Nunc International , Rochester , NY ) , that were pre-coated with lysates from axenic parasites , infected cells or uninfected cells . A Goat anti-mouse-HRP secondary antibody was applied . Plates were developed using HRP substrate solutions ( BD Biosciences , San Jose , CA ) and plates were read at 450 nm on a Powerwave 200 spectrophotometer ( Bio-Tek , Winooski , VT ) . The genomic expression library was grown on LB medium containing kanamycin ( 50 mg ml−1 ) for 12–14 H at 37°C to generate plates containing approximately 200–500 colonies . Each plate was replicated using nitrocellulose disks that were then placed onto a LB plate containing kanamycin and IPTG ( 1 mM ) and incubated for 5 H at 37°C to induce expression of cloned open reading frames ( ORFs ) . The colonies were exposed to chloroform vapors for 20 min , then overlaid with nitrocellulose membranes . The membranes were incubated in PBS , pH 7 . 2 , containing 0 . 5% Tween-20 ( PBS-Tween ) and 5% non-fat skim-milk . They were reacted with the adsorbed sera at a 1∶300 dilution in PBS-Tween for 1 H , then with peroxidase-conjugated goat anti-mouse immunoglobulin at a 1∶3000 dilution for 1 H . Reactive clones were detected by enhanced chemiluminescence ( ECL kit , Amersham Pharmacia ) and exposure to Hyperfilm ( Amersham Pharmacia ) . To confirm the immunoreactivity of the clones identified above , several colonies from the master plate in the apparent vicinity of the signal were streaked for single colony isolation onto LB plates . After overnight incubation , three single colonies isolated from each original colony were resuspended in fresh liquid LB medium containing kanamycin . Aliquots were blotted onto solid LB medium containing kanamycin and IPTG . Following the induction of recombinant protein expression , the blotted cells were lysed and re-screened using the same procedures as in the first round of screening . Two negative controls were included on each plate: pET30b/BL21 ( DE3 ) with no cloned insert , and a random clone that contained an insert but was non-reactive with sera . PCR primers to amplify the gene transcripts of the antigens identified with CMAT were designed using the sequence in the clones or the sequence of the closest homolog available in the sequence database ( www . genedb . org ) as well as the 5′ splice leader . For example , for the IVI-59 , the L . infantum ortholog ( LinJ31_V3 . 1500 ) was used to design primers to amplify a transcript fragment corresponding to nucleotide positions +376 to +2113 of the L . infantum sequence . The PCR primers used were FB535 5′-CCATGGACCTGCCCACCGTCACATTC-3′ and FB536 5′-AGCTTGCCACGCTGCCACCTCTTTGAG-3′ . The PCR product was electrophoresed in a 1% agarose gel and a fragment at the expected size location ( 1771 bp ) was observed . The PCR fragment was excised from the gel , purified using a gel extraction kit ( Qiagen ) and cloned into a pET30 expression vector ( Novagen ) using restriction sites incorporated into the primers . RAW 264 . 7 macrophages plated in 100 mm dishes were infected with L . donovani or L . pifanoi promastigotes . At various times , parasites were removed , cell monolayers washed twice with PBS and the infected macrophages were lysed directly in the plate with lysis buffer ( 25 mM HEPES , 150 mM NaCl , 1% Triton X-100 , 1x protease inhibitor pill ( 1 pill/10 mL ) ( Roche Applied Science , Indianapolis , IN ) , 20 ul/mL Phosphatase Inhibitor Cocktail 1 ( Sigma-Aldrich , St . Louis , MO ) , 20 ul/mL Phosphatase Inhibitor Cocktail 2 ( Sigma ) , 50 uM NaVO4 ) . Protein containing lysates were then cleared of cellular debris by centrifugation and protein concentration was determined using a Bradford Protein Assay ( BioRad , Hercules , CA ) . SDS-PAGE was run on 50 µg of each sample on 12% polyacrylamide gels . Proteins were transferred to Immobilon P membrane ( Millipore , Billerica , MA ) , the membranes were incubated with milk and then probed with primary antibody . After removal of primary antibodies and washing , membranes were incubated in the appropriate secondary antibodies conjugated to horse-radish peroxidase . Washed blots were incubated with chemiluminescence ( ECL , Amersham ) reagents . Antibody reactivity was visualized by exposure of blots to x-ray film . Some blots were stripped by incubation in 62 . 5 mM Tris HCl pH 6 . 8 supplemented with 20 mM 2-mercaptoethanol and 2% SDS , for 30 minutes at 50°C . The blots were then re-probed with other antibodies including the JLA20 the anti-actin antibody and E7 the anti tubulin antibody ( both of these antibodies were obtained from the Developmental Studies Hybridoma Bank ( University of Iowa ) . Infected cells on coverslips were fixed with 2% paraformaldehyde and processed as described previously [25] . Incubation with primary antibodies was performed in binding buffer supplemented with 0 . 05% saponin . Coverslips were incubated with the appropriate secondary antibodies into which the nucleic acid dye 4′ , 6-diamidino-2 phenylindole dihydrochloride ( DAPI ) had been added . Coverslips were mounted on glass slides with ProLong antifade ( Molecular Probes ) . They were examined through a 63X oil immersion lens on a Zeiss axiovert 200 M integrated into a spinning disc confocal microscope technology from PerkinElmer ( Waltham , MA ) controlled by the volocity software . Z stack of optical sections spanning the entire cell were captured and then combined using the extended focus feature in the volocity software producing a 3D image . Two methods of cell preparation were used . In the first , cells infected for 72 H were fixed for 10 min at room temp then for 20 minutes at 4°C in a mixture of 4% formaldehyde and 1% glutaraldehyde in 0 . 1 M cacodylate buffer ( pH 7 . 2 ) . The pellet , embedded in low-gelling-point agarose to facilitate handling , was dehydrated in an ethanol series , then embedded in LR White resin ( Electron Microscopy Science , Fort Washington , PA ) and polymerized at 50°C . Sections were cut with a diamond knife on an RMC MT-6000-XL ultramicrotome and collected on Formvar-coated nickel grids . In the second method , 72 H-infected cultures were scrapped into a small volume of 0 . 15 M sucrose in RPMI . The cells were pelleted in 1 . 5 ml vials and they were rapidly frozen in a Baltec HPM 010 high-pressure freezer ( Boeckeler Instruments , Tucson , AZ ) . Freeze-substitution was carried out 5 days at −80°C in anhydrous acetone containing 2% OsO4/0 . 5% uranyl acetate . The temperature of the freeze-substitution medium was then increased from −90°C to −60°C over 12 H , and the samples were washed with anhydrous acetone three times at −60°C . Lowicryl HM20 ( Electron Microscopy Sciences ) resin embedding was performed by a stepwise increase in resin concentration from 0 , 33 , 66 , to 100% over 48 H at −60°C . The samples were then washed three times with 100% resin and polymerized under UV light for 24 H at −60°C . Once completely polymerized , the samples were slowly warmed from −60 to 20°C over a 4 H period as described in [26] . Thick sections ( 80 to approx . 120 nm thick ) of Lowicryl HM20 embedded cells were mounted on gold slot grids coated with formvar ( Electron Microscopy Sciences , Hatfield , PA ) . Grids were blocked with 1% dry milk in phosphate-buffered saline ( PBS , pH 7 . 2 ) , treated overnight at 4°C with primary antibody or normal mouse serum ( control ) diluted 1∶10 in PBS ( chemically fixed samples ) or 1∶25 ( cryofixed samples ) , washed with high salt Tris/Tween buffer ( 0 . 5 M NaCl; 0 . 02 M Tris , pH 8 . 0; 0 . 1% Tween-20 ) and reacted for 1 H with secondary antibody conjugated to 12 nm or 18 nm gold ( goat anti-mouse IgG; Jackson ImmunoResearch Laboratories , West Grove , Pa . ) diluted 1∶30 in PBS . In experiments with antisera to cTXNPx ( Rat ) a secondary antibody conjugated to 5 nm gold was used . Some sections were exposed to gold-labeled secondary antibody alone as an additional control . After washing , grids were floated on Trump's fixative for 10 min and washed with deionized water . Sections were then post-stained with 0 . 5% uranyl acetate and lead citrate and were examined with a Zeiss EM-10CA transmission electron microscope Immuno-EM labeling was performed as described . Images were captured on Hitachi 700 TEM . Images were opened in ImageJ . The area of the parasite , PV lumen , cytosol and nucleus were measured and normalized using the measure bar on the image . Gold particles were counted in each compartment and normalized by dividing by the area of the compartment ( µm2 ) . N = 10 . RAW264 . 7 cells were cultured as described and infected with L . pifanoi at ratio of 1∶10 . Cells were harvested at 48 and 72 H . Cells were washed with cold PBS three times . They were then lysed in a nuclear extraction buffer ( 10 mM Hepes , 10 mM KCl , 1% Triton X 100 , 0 . 1 mM Na3VO4 , 2 mM MgCl2 , 0 . 5 mM DTT , 1 mM PMSF ) and laid over 20% Ficoll-Paque ( Pharmacia ) as described in [27] . The lysate was centrifuged at 13 , 000 rpm to separate the cytoplasm from the nuclei . The supernatant was removed and saved as crude cytosolic fraction . The nuclear pellet was washed 3 times using the nuclear extraction buffer . Aliquots of this fraction that were monitored by light microscopy did not reveal the presence of intact parasites . The nuclear pellet was then lysed in RIPA buffer to release nuclear proteins . The Protein containing lysates were then cleared of cellular debris by centrifugation and protein concentration was determined using a Bradford Protein Assay ( BioRad , Hercules , CA ) . SDS-PAGE was run on 50 µg of each sample on 12% polyacrylamide gels . Proteins were transferred to Immobilon P membrane ( Millipore , Billerica , MA ) .
We elected to focus our studies on two Leishmania species: L . mexicana pifanoi ( MHOM/VE/57/LL1 ) a member of the L . mexicana complex and L . donovani ( MHOM/SD/62/1S-CL2 ) . The rationale for selecting these parasites included the following: 1 ) Experimental conditions for the transformation of these parasites lines from the promastigote form to the amastigote form and their maintenance as amastigotes is well established [22] , [28]–[34] . Obtaining sufficiently adsorbed antibodies is a critical step in the CMAT method , so it was important to have parasite lines with stable growth in the amastigote stage under reproducible conditions . 2 ) These two parasite species reside in morphologically distinct PVs . L . pifanoi parasites reside in communal PVs that undergo progressive distention over the length of the infection period . In contrast , L . donovani parasites reside in tight PVs from where daughter progeny segregate into new PVs . Since we did not know which of these two intracellular environments would represent a greater change in the parasite's environment that would induce differential gene expression , it was prudent to use both parasites in our analyses . To obtain serum that was preferentially reactive to infected cells , lysates of RAW 264 . 7 macrophages infected for 24 H with axenic amastigote forms of either Leishmania species were prepared and used to immunize BALB/c mice . Immune serum was collected and adsorbed extensively with whole amastigotes , amastigote lysates , heat-killed amastigotes , and lysates of RAW 264 . 7 cells as described in the Materials and Methods section . The resulting adsorbed serum had minimal reactivity with lysates from axenic parasites , but retained reactivity to lysates from infected cells . A representative ELISA showing the reactivity of pre-adsorbed and adsorbed sera to axenic L . pifanoi ( LL1 ) lysates and infected cell lysate is shown in Figure 1 . Even at a 1∶50 dilution of the adsorbed antisera , these sera had no reactivity to lysates prepared from axenic parasites even though they retained their reactivity to lysates prepared from infected cells . Similar results were obtained with L . donovani reactive serum ( not shown ) . This implied that antibodies reactive to molecules expressed by axenic parasites were depleted during the adsorption protocol . However , the possibility that the adsorption of molecules expressed at low levels by axenic parasites was incomplete cannot be ruled out . The CMAT is an immunoscreen and its success is dependent on the quality of the adsorbed immune serum as judged by its titer and the diversity of the antigenic determinants to which the serum is reactive . In the future , sera with more diverse immune reactivity might be obtained by immunizing outbred mice species or rabbits . Alternatively , pooled convalescence sera from human subjects would provide a rich serum source with diverse antigenic reactivity . The adsorbed immune sera were used to identify the new antigenic parasite molecules that they were reactive to within infected cells . To achieve this , these sera were used to screen inducible expression libraries in E . coli BL21 ( DE3 ) cells containing Leishmania genomic inserts from 0 . 5 to 2 . 0 Kbp cloned in pET30abc ( Novagen ) . The use of genomic expression libraries is feasible in leishmaniasis because only a few parasite genes have been shown to have introns . After a primary screen of 20 , 000 clones from the L . pifanoi library and 20 , 000 clones from the L . donovani library , we obtained 27 clones reactive to the L . pifanoi serum and 30 clones reactive to the L . donovani serum . Although these library screens did not represent a total coverage of the genome , we decided to proceed in order to evaluate the characteristics of the molecules that were being identified by this approach . Twenty total clones remained robustly positive after tertiary screens . Four clones reactive to L . pifanoi and seven reactive to L . donovani were randomly selected for further analyses . The inserts in the clones were sequenced and the sequences were used to BLAST ( Basic Local Alignment Search Tool ) search Leishmania genomes in the sequence database www . genedb . org . Table 1 shows the most likely homologs of the in vivo induced genes ( IVI ) identified by CMAT . Six of the 10 IVI genes selected were homologs of genes that encode hypothetical proteins for which the Leishmania genes have no known homologs in organisms other than kinetoplastids . Three of the remaining genes belonged to gene families that encoded parasite molecules that had previously been shown to be preferentially expressed in Leishmania amastigote forms . The tryparedoxin peroxidase gene had been characterized in promastigote and amastigote forms . There was only one instance ( IVI-7 . 1 ) , in which we were not able to associate a gene from the sequenced genomes to the sequence in the CMAT clone even though a homologous sequence was found in the L . infantum genome . A concern with the selection of the parasite lines used in this study was that genes in organisms such as L . pifanoi and L . donovani , whose genomes had not been completely sequenced , might not have homologs in the genomes that had been sequenced . This concern was somewhat mitigated by the study in which comparison of a few selected open reading frames ( ORFs ) of L . mexicana were found to have a 89 . 5% identity to L . major ORFs [35] . However , a study by Peacock and colleagues [36] , in which the genomes of L . braziliensis , L . infantum and L . major were compared , identified 78 genes that are restricted to either genome and observed an additional 200 sequence differences . Whether any of such unique genes would have been detected in our screen is unknown . With the exception of IVI-7 . 1 noted above , all the IVI genes identified from the L . pifanoi and L . donovani expression libraries were found to have homologs in the L . major , L . infantum and L . braziliensis sequenced genomes . The sequence of the L . mexicana and L . donovani genomes are now available . An intracellular variant of Typaredoxin peroxidase . The BLAST search of the sequenced genomes with the IVI-16 sequence revealed that this CMAT clone contained a L . pifanoi homolog of tryparedoxin peroxidases ( TXNPx ) . The clone contained the 3′ region of the TXNPx gene and a segment of the contiguous gene on the chromosome annotated as the developmentally regulated gene ( DRG ) . Induction of IVI-16 expression with IPTG produced a 15 kDa recombinant protein that included the c-terminus of the TXNPx gene and vector sequences . A stop codon between the TXNPx and DRG sequences prevented expression of the DRG sequence in the clone . Using primers designed from the sequence in the IVI-16 clone and the Leishmania splice leader , we obtained the entire coding sequence of the gene in the CMAT clone by amplification of cDNA prepared from L . pifanoi-infected cells . A phylogentic analysis of the relatedness of this sequence with other available TXNPx sequences showed that this gene is most closely related to AAX47428 , a gene from L . amazonensis , which like L . pifanoi is a member of the L . mexicana complex ( Figure S1 ) . An alignment of the predicted protein sequence of the new IVI-16 gene product with the available sequences of TXNPx in the GeneDB genome database is shown in Figure 2 . In addition to amino acid sequence differences in the C-terminus of these molecules , some TXNPx molecules including the IVI-16 gene cloned here are nine amino acids shorter than other TXNPx sequences . This is consistent with a previous study that demonstrated the amastigote variant ( P1 ) of L . chagasi peroxidoxin is shorter by 9 amino acids than two other variants ( P2 and P3 ) preferentially expressed in L . chagasi promastigotes [37] . This suggested that the gene identified by CMAT is amastigote specific . This cloned variant of the L . pifanoi TXNPx gene is hereto forth referred to as IVI-16/TXNPx . We also noted that there are additional amino acid differences throughout the length of the protein that might result in antigenic differences between this protein and other TXNPx expressed by Leishmania parasites . Tryparedoxin peroxidases ( also called peroxidoxins or peroxiredoxins ) belong to a gene family that has been characterized extensively in Leishmania [37] , [38] , [39] and in other organisms as well [40] . These molecules are antioxidants that detoxify peroxides , hydroxyl radicals and peroxynitrite by several mechanisms including oxidoreduction . These molecules are vital to the parasite since it is believed that both in the insect vector and in the mammalian host Leishmania encounter a range of toxic oxidative environments that it must harness to survive . Moreover , recent reports have implicated TXNPx in parasite dissemination and in light of the inability to generate parasites that lack this gene , it was suggested that TXNPx are necessary for parasite survival [38] . TXNPx variants have been described that localize to discrete cellular compartments such as the mitochondrion and the cytosol within the parasite [37] , [38] . As discussed above , the CMAT screen resulted in the cloning of the L . pifanoi IVI-16/TXNPx . Given the relatedness of the cloned gene to other TXNPx genes , it was intriguing that antibodies to this molecule were not adsorbed out after incubations with axenic parasite lysates . The possibility that the adsorption protocol implemented was incomplete could not be ruled out; alternatively , it was likely that the IVI-16/TXNPx molecule identified is a variant that is antigenically different from other TXNPx variants expressed in the parasite . Several pieces of evidence support the latter statement . Western blot analysis of lysates obtained from cultured organisms , as well as lysates from infected cells and uninfected macrophages , probed with antiserum raised with the IVI-16/TXNPx recombinant protein showed that the antiserum was not reactive to lysates of uninfected macrophages ( lane M ) but exhibited strong and specific reactivity to a molecule of ∼20 kDa in lysates obtained from macrophages infected for 24 , 48 or 72 H with L . pifanoi promastigotes ( Figure 3 ) . Promastigote stage parasites transform into the amastigote form in 10–24 H after entry into macrophages [41] . No comparable reactivity was observed to lysates of cultured promastigotes alone ( P ) or axenic amastigotes ( A ) . The same lysates were tested for their reactivity to P8 antiserum . P8 is a molecular complex that is expressed in stationary stage promastigotes and axenic amastigotes of L . pifanoi [42] . The complex was indeed identified within the parasite lysates as well as within infected cells ( Figure 3 ) . The anti-P8 antiserum was also non-reactive with lysates from uninfected macrophages and actin levels in lysates from uninfected and infected macrophages confirmed that the gel was loaded equally ( Figure 3 ) . Northern blots probed with the IV1-16/TXNPx sequence under stringent conditions revealed the presence of transcripts of the endogenous IV1-16/TXNPx gene in axenic promastigotes and amastigotes as well as infected cells , which is consistent with reports of Leishmania genes being constitutively transcribed ( not shown ) . Additional evidence that the IV1-16/TXNPx molecule identified by the CMAT screen is antigenically distinct from a cytosolic variant of TXNPx ( cTXNPx ) expressed by these parasites was obtained in protein depletion experiments . Here , lysates obtained from infected macrophages were incubated with the anti-IVI-16/TXNPx antiserum . The endogenous molecule in the lysate that is reactive to this antiserum was depleted by adding protein G to the lysate . The supernatant fluid from this depletion experiment was then analyzed in Western blot assays for reactivity to the anti-IVI-16/TXNPx antiserum and to an antiserum that is reactive to the cTXNPx [39] . The incubation of infected cell lysates with the anti-IVI-16/TXNPx depleted the endogenous molecules recognized by that antiserum while the molecule that is recognized by the antiserum to cTXNPx was not depleted ( Figure 3B ) . Taken together , these results show that IV1-16/TXNPx molecule , which is expressed within infected cells , is antigenically different from the cTXNPx that is also expressed in infected cells as well as by axenically cultured organisms . Six of the CMAT clones were homologs of genes annotated as encoding hypothetical proteins in the sequenced genomes ( Table 1 ) . This proportion of hypothetical proteins identified by the CMAT screen should be expected since only a third of the genes in the Leishmania genome have known homologs [43] . Application of sequence analyses tools on the homologs of the genes that were identified revealed that two of these genes have domains that might be suggestive of their function . The LmjF25 . 0450 homolog identified by IVI-4 was found to contain a GTP-binding domain , which suggests that this molecule might play a role in signal transduction [44] . An NCBI PSI-blast of the LmjF30 . 2390 gene identified by IVI-18 revealed that in addition to its relatedness to serine peptidases , it contains both a leucine rich repeat region ( LRR ) and hemopexin repeats . Over 100 proteins in L . major have been estimated to contain LRR sequences , which mediate protein-protein interactions . Another member of the LRR superfamily , LinJ34 . 0570 , was recently described and implicated in antimony resistance [45] . Hemopexin repeats in proteins have been shown to mediate interactions with heme [46] . This is of relevance to Leishmania biology since it has been shown previously that Leishmania infections modulate heme degradation in infected cells , which controls the capacity of the infected cell to elaborate responses such as superoxide production that are dependent on heme availability [25] . It is also of interest that a molecule with heme binding activity was recently described on the surface of L . infantum amastigotes [47] . Future studies on these molecules should be illuminating . Antisera were generated to the recombinant molecules encoded by the original library clones in pET30 . The antisera were first confirmed for their reactivity to the recombinant antigens ( not shown ) before they were used to detect the endogenous molecules expressed in infected cells by Western blot assays , in immunofluorescence assays and immune-electron microscopy . Only antisera raised to the recombinant products of IVI-4 , IVI-18 and IVI-59 , which are the homologs of LmjF25 . 0450 , LmjF30 . 2390 and LinJ31_V3 . 1500 respectively , reacted robustly to cells infected for 24 to 72 H . These antisera were reactive to infections with either the L . pifanoi or L . donovani parasite lines used to initiate the studies . Consequently , only analyses performed on L . pifanoi infected cells are presented . The antiserum to IVI-4 was poorly reactive in Western blot assays but exhibited specific reactivity to a parasite structure within infected cells ( see below ) . Representative Western blot analyses to evaluate the expression of the endogenous genes identified by IVI-18 and IVI-59 are shown in Figure 4 . In Figure 4A the reactivity of antiserum to IVI-18 is shown . This antiserum was not reactive to lysates of uninfected macrophages ( lane M ) . In lysates prepared from macrophages infected for 24 to 72 H , a doublet of ∼95 kDa is recognized by this antiserum . In contrast , no molecule of comparable molecular weight is recognized in lysates prepared from stationary promastigote cultures or lysates from axenic amastigotes . In Figure 4B , the reactivity of the antiserum to IVI-59 is shown . This antiserum too was not reactive to lysates from uninfected macrophages ( lane M ) . The IVI-59 antiserum was reactive to a molecule in infected cells of ∼125 kDa . Other bands in infected samples were most likely the result of protein degradation since the intensity of these bands varied with each lysate preparation . This antiserum had limited and inconsistent reactivity to lysates from axenic amastigotes and lysates obtained from stationary stage promastigotes . Moreover , any reactivity that was observed to promastigote or amastigote preparations was at a higher molecular weight . Together these results show that the L . pifanoi homolog of LmjF30 . 2390 and LinJ31_V3 . 1500 are preferentially expressed within infected cells . The apparent reduction in the levels of the endogenous proteins at the 72 H point was most likely due to the loss of infected cells that detach easily from the culture dish . The three remaining molecules that were identified by the library screens were found to be homologs of previously annotated genes in the sequenced genomes . IVI-64 had similarity to amastins , which are a family of proteins that have been shown to be differentially expressed in Leishmania parasites [48] , [49] . Alignment of the IVI- 64 sequence with sequences of amastin genes showed that it had highest similarity to LmjF34 . 1080 ( Figure S2 ) , which is an amastin variant predicted to be expressed on the parasite surface of amastigotes forms [46] . Knowledge of the conditions that result in the differential expression of individual members of the amastin gene family is incomplete . However , it has been shown that amastins gene expression , analyzed at the mRNA level , is regulated by the pH of their environment and that the highest accumulation of amastins is evident after 5–7 days of amastigote culture [48] , [49] . It is likely that similar to the situation with IVI-16/TXNPx , the immunization with lysates from infected cells resulted in the generation of antisera to an antigenically distinct variant of amastins that is preferentially expressed within infected cells . The IVI-63 had similarity to the am3 gene ( encoding for Ama1 protein ) , which was previously identified from a differentially expressed cDNA library and shown to be amastigote specific [50] . Although the time course of appearance of the endogenous protein was not evaluated in that study , transgenic parasites with an episomal copy of the gene , were shown to express this protein on the surface of parasites within infected cells . Further observations on this molecule are described below . Finally , BLAST of the NCBI database with the IVI-62 sequence showed that it had 99% similarity to the L . infantum cysteine proteinase b gene ( AJ420286 ) . Alignments of the IVI-62 sequence with the sequence of other cysteine proteinases ( not shown ) have provided additional evidence that it is indeed a cysteine proteinase . Immunofluorescence assays and immuno-electron microscopic analyses were performed to determine the distribution within infected cells of the endogenous molecules that were identified by the CMAT screen . The distribution of the endogenous IVI-16/TXNPx molecule within infected cells was determined . The IVI-16/TXNPx antiserum labeled the parasite within infected cells ( Figure 5 , [thin white arrow points to a representative parasite] ) . There was also labeling of the PV in regions where there were no parasites . In addition , the antiserum labeled structures in the cytosol outside the PV ( thick white arrows ) . The pattern of labeling of endogenous IVI-16/TXNPx was compared to the pattern of labeling of previously described antisera to the cTXNPx [39] and the mitochondrial TXNPx ( mTXNPx ) [38] . Both the antiserum to cTXNPx and mTXNPx exhibited specific and restricted labeling of the parasite within infected cells with patterns that were consistent with cytoplasmic staining and mitochondrial staining respectively; they did not label any structures outside the parasite . To obtain greater resolution of the structures that were reactive to the IVI-16/TXNPx antiserum , immuno-labeling of infected cell sections on grids that were processed for EM analysis was performed . Two protocols for sample preparation for EM analyses were used . LR white embedding of chemically fixed cells permitted better resolution of membranous structures as compared to cryofixation of samples . However , all the antisera that were used in this study labeled sections prepared with the latter protocol with greater sensitivity . EM images using both of these protocols are presented in Figure 6 . In Figure 6A a representative example of the pattern of labeling observed when the IVI-16/TXNPx antiserum was applied on samples obtained by cryofixation is shown . In these samples the antiserum labeled the parasite , the PV lumen and the cell cytosol outside the PV ( black arrows ) . The pattern of labeling of consecutive sections with NMS was ascertained for comparison . A representative image of a cell incubated with NMS and processed identically to those incubated with immune serum is shown in Figure 6A ( arrows in this image point to gold particles ) . To quantify the proportion of gold labeling that was attributable to the reactivity of the IVI-16/TXNPx antiserum , gold particles in EM sections were enumerated . Figure 6B shows that in sections incubated with the IVI-16/TXNPx antiserum there were approximately33 gold particles/µm2 on the parasite , 2 particles/µm2 in the PV lumen , 1 particle/µm2 in the cytosol and 3 . 5 particles/µm2 in the nucleus . In contrast , there were 1 , 0 . 14 , 0 . 35 and 1 . 65 gold particles/µm2 on the parasite , PV , cytosol and nucleus , respectively , when NMS was applied on these sections . With the exception of the nucleus , the difference between the concentration of gold particles in within cellular compartments of sections incubated with IVI-16/TXNPx antiserum was significantly ( p<0 . 05 t-test ) higher than for particles on sections incubated with NMS . Analysis of samples processed after fixation and LR white embedding also showed that the IVI-16/TXNPx antiserum labeled the parasite with a pattern that is best described as cytosolic . In addition , there was evidence of labeling of the PV lumen and vesicles in the cell cytosol outside the PV ( Figure 6C ) [white arrows point to gold particles and vesicular structures] . The vesicular structures in the cytosol , ( some of which are labeled ) had a ‘fuzzy’ coat that is reminiscent of coated vesicles . Labeling of such vesicles was observed only with the antiserum to IVI-16/TXNPx . The pattern of labeling of uninfected cells with this antiserum was similar to the pattern of labeling of infected cells with normal mouse serum ( NMS ) . To address the possibility that the observed labeling in the PV and the infected cell cytosol might be derived from dead parasites , fresh grids containing sections of infected cells were double labeled with the antiserum to the cTXNPx and IVI-16/TXNPx antiserum . A representative image showing the pattern of a double labeled cell is shown in Figure 6D . Reactivity of the cTXNPx was visualized with 5 nm gold particles [thin arrows] . The cTXNPx antiserum resulted in robust labeling of the parasite forming a pattern of labeling that is consistent with a cytoplasmic distribution of the endogenous molecule . The antibody to cTXNPx exhibited minimal reactivity outside the parasite . Labeling with the IVI-16/TXNPx antiserum also resulted in a pattern on the parasite that was consistent with a cytoplasmic distribution pattern . However , it is evident that the IVI-16/TXNPx antiserum labels structures in the PV lumen outside the parasite as well as the host cell cytosol ( thick black arrows ) . Gold particles from such double labeling experiments were enumerated ( Figure 6E ) . There were significantly fewer 5 nm gold particles ( cTXNPx ) in the PV lumen , the host cell cytosol and nucleus as compared to large gold particles ( IVI-16/TXNPx ) . A surprising observation , however , was that there was less 18 nm gold and 5 nm gold labeling on the parasites in the double labeling experiments than what was observed in the single labeling experiments . This suggested that both antibodies competed for antibody binding sites but only on molecules that were expressed within the parasite . Nonetheless , these results provide additional evidence that the endogenous IVI-16/TXNPx variant is antigenically different from the cytosolic variant previously described . Furthermore , the endogenous IVI-TXNPx molecule appears to traffic out of the parasite into the PV and beyond the PV into the host cell cytosol in coated vesicles . The double labeling experiments argue against the likelihood that labeling in the PV lumen and within the infected cell cytosol was derived from dead parasite material . The distribution of the endogenous molecules that are reactive to antisera to IVI-18 and IVI-59 was also investigated in parallel experiments . Recall that these antisera too were raised to recombinant molecules and that they exhibited no reactivity to macrophage molecules in non-infected cells . Immunofluorescence assays of infected cells with antiserum to IVI-59 showed that this molecule labeled the parasites within PVs ( thin white arrow points to representative parasite ) ( Figure 7 ) . In addition , there was labeling of the PV lumen , PV membrane and structures in the host cell cytosol ( bold white arrows ) . We proceeded to perform immuno-EM analysis to assess further , the reactivity of the IVI-59 antiserum . Sections obtained by LR white embedding of chemically fixed cells showed that the endogenous molecule identified by this antiserum is localized primarily to the parasite surface ( Figure 8A [arrows point to examples of gold particles ) . Unlike the observation made with antiserum to IVI-16/TXNPx that labeled what appeared to be coated vesicles in the host cell cytosol , the IVI-59 antiserum labeled vesicles of various sizes that had no other distinguishing characteristics in the host cell cytosol . We next analyzed sections that were processed after cryofixation . The distribution pattern of the IVI-59 molecule on the parasite was as observed in the chemically fixed and LR White embedded sections . Additional observations were made with the cryofixed samples . One of the more intriguing observations was the presence of gold particles in membrane enclosed compartments within the parasite and also on vesicular structures on the parasite surface ( Figure 8B [thin black arrows] ) . The labeled vesicular structures on the parasite surface appeared to have ‘bubbled’ from the parasite These vesicular structures on the parasite surface are reminiscent of exosomes recently described by Silverman and colleagues [51] . However , these vesicles did not appear to have been released from the flagellar pocket . Another interesting observation with this antiserum was the apparent transfer of the endogenous IVI-59 molecule from parasites in the PV to host cell vesicles ( Figure 8C [blue arrows point to gold particles on the parasite surface , at the parasite-macrophage interface and in vesicular structures within the host cell] ) . The recipient vesicle in the host cytosol had similar characteristics to other vesicles of various sizes with no other outstanding characteristics that were labeled with gold particles in the host cell cytosol . Finally , immuno-labeling of these sections also exhibited reactivity of the host cell nucleus . Gold particles on the parasite , the PV lumen , the host cell cytosol and the host cell nucleus were enumerated and compared to the distribution of gold particles on sections that were incubated with NMS . Figure 8D shows that there was significantly more gold particles in all the compartments analyzed when the sections were incubated with the antiserum to IVI-59 . Double labeling experiments with the IVI-59 antiserum and the antiserum to cTXNPx were also performed; a representative image of a double labeled cell is shown in Figure 8E . The cTXNPx exhibits robust labeling of the parasite ( black arrows ) . In contrast the IVI-59 antiserum primarily labels the parasite surface ( bold white arrows ) . The concentration of gold particles on these sections was determined and plotted ( figure 8F ) . There was considerable labeling of the parasite by the antiserum to cTXNPx but minimal reactivity elsewhere . Interestingly there was no apparent interference between the cTXNPx labeling and the labeling with the IVI-59 antiserum as was noted with the IVI-16/TXNPx antiserum . Taken together , analysis of the distribution of the endogenous molecule ( homolog of LinJ31_V3 . 1500 ) that is reactive with IVI-59 antiserum showed that this molecule is expressed primarily on the parasite surface . However , it is released into infected cells and traffics outside of the PV to the host cell nucleus . Here too , the double labeling experiments suggested that the extra-parasite labeling of the IVI-59 antiserum is unlikely to be the result of trafficking of molecules from dead parasites . The distribution of the IVI-18 molecules was also analyzed in parallel experiments . Figure 7 shows a representative image of an infected cell captured after immunofluorescence staining with antiserum to IVI-18 . There is specific staining of parasites within infected cells . We proceeded to analyze samples processed for immuno-EM analyses . There was no difference of interest between samples processed by the chemical fixation and LR white embedding protocol and the cryofixed samples . A representative image from cryofixed samples shows the distribution pattern of the endogenous molecule recognized by the IVI-18 antiserum ( Figure 9A [black arrows point out representative gold particles] ) . This molecule appears to be expressed primarily on the parasite surface . Gold particles on the parasite , PV lumen , host cell cytosol and host cell nucleus were enumerated and plotted . When sections were incubated with the IVI-18 antiserum , the number of gold particles on parasites was significantly higher than that obtained with sections incubated with NMS ( Figure 9B ) . There were also more gold particles in the other compartments analyzed as compared to incubation with NMS . It is noteworthy that the overall concentration of gold particles in sections incubated with the IVI-18 antiserum was less than what was obtained with the IVI-59 antiserum , for example . Taken together , these observations too show that the homolog of LmjF30 . 2390 that is recognized by the antiserum to IVI-18 is primarily localized to the parasite surface but that it also gains access to the host cell cytosol and nucleus . The reactivity of the antiserum raised to the IVI-4 was unique as compared to the other CMAT molecules . Although there was strong reactivity to parasites within PVs there was minimal reactivity of this antiserum elsewhere in the infected cell ( Figure 7 ) . This suggested that the homolog of LmjF25 . 0450 identified by IVI-4 is neither released into the PV nor secreted into the host cell cytosol . Representative confocal images of the reactivity of antisera to IVI-62 , IVI-63 and IVI- 64 are included in the supplemental figures ( Figure S3 ) . Both IVI-63 and IVI- 64 appear to be localized to the surface of parasites within PVs . These observations confirm previous reports that had shown that the AMA molecule ( homolog of IVI-63 ) and some amastin variants ( IVI-64 ) are expressed on the parasite's surface [49] . Table 2 shows a summary of the distribution of the endogenous molecules identified in this study . The localization studies of the endogenous IVI molecules described above suggested that some of those parasite molecules might traffic to the host cell nucleus . To determine whether endogenous IVI molecules are localized in the nucleus of infected cells , a subcellular fractionation protocol to isolate the nucleus of infected cells was implemented . The resulting nuclear and cytosolic fractions were analyzed by Western blot analysis for their reactivity with antisera to IVI-16/TXNPx , IVI-18 and IVI-59 . After 48 and 72 H infections , the IVI-16/TXNPx antiserum is exclusively reactive with the cytosolic fraction ( Figure 10 ) . Similarly , the antiserum to IVI-18 is mostly reactive with the cytosolic fraction with occasional faint reactivity with the nuclear fraction . In contrast , the IVI-59 antiserum is reactive with both the nuclear and cytosolic fractions . Its reactivity with the nuclear fraction is with a molecule that is at the appropriate molecular weight . The quality of the subcellular fractionation was monitored with the reactivity of DE6 , an antibody that is specific for a 47/51 kDa nucleolar protein [52] . Taken together , these observations complement the results obtained by counting gold particles on EM sections that had found that the labeling of the nucleus with the IVI-59 antiserum was significantly more than that obtained with NMS . In contrast although there was more gold labeling of the host nucleus with antisera to IVI-16/TXNPx and IVI-18 , the difference was not significantly different from the labeling with NMS . An important caveat in the interpretation of these results though is that the subcellular fractionation protocol as implemented was stringent and could therefore exclude some molecules that traffic to the nucleus under physiological conditions . Prior to the observation reported here of Leishmania parasite molecules trafficking in the infected cell , elongation factor 1α ( EF-1α ) from Leishmania had been shown to localize to the host cell cytosol where it targets SHP-1 [53] , [54] . EF-1α was identified within macrophages after 16 hours of infection by promastigotes , however the mechanism by which EF-1α accesses the host cell cytoplasm has not been described . Several lines of evidence have also suggested that cysteine proteinases too , must gain access to the infected cell cytosol where they target components of the NFκB signaling complex [55] . This led Mottram and colleagues to propose that cysteine proteinases are most likely transported out of the PV in vesicles which rupture in the cytosol [56] . In this study , IVI-62 ( a cysteine proteinase homolog ) was identified by the CMAT screen and found to gain access to the cell cytosol . In addition to IVI-62 , several lines of evidence were obtained that showed that other Leishmania parasite molecules are released beyond the PV as well . It is noteworthy that the presence or absence of a signal sequence was not predictive of the trafficking scheme of parasite molecules in the infected cell . This observation is in agreement with a recent study in which parasites molecules that are secreted into culture medium ( secretome ) were analyzed [57]; in that study , it was concluded that most of the molecules identified did not appear to be secreted through the classical secretion pathway . The authors discussed that Leishmania might possess several secretion pathways in addition to secretion through the flagellar pocket . In this study , the vesicular structures that were reactive with the IVI-59 antiserum might be a part of one of those secretion pathways . Further studies employing reagents to the parasite molecules identified in this study should provide greater insight into protein secretion by Leishmania parasites within infected cells . The protein products of the CMAT genes analyzed thus far were detected in infected cells after many hours ( days ) of infection , however , their transcripts were detected in parasites that were cultured axenically ( not shown ) . Studies on LIT1 , the ZIP family iron transporter had also found that although the transcripts of LIT1 are found in promastigotes and axenic amastigotes , the protein product is detected only in parasites within PVs after several days of infection [12] . Together , these observations suggest that of the estimated 8300 Leishmania proteins there is a subset of genes for which the protein products are synthesized only in the intracellular environment . This subset might also include parasite molecules that are preferentially expressed at discrete times in the infected host . It is not known how many molecules would fall into this subset . Such molecules are of great interest because they might mediate the pathogenic mechanisms of the parasite and represent biomarkers of disease progression in the infected host . CMAT , like IVIAT , has limitations that should be acknowledged . It is an immunoscreen so only molecules that are immunogenic will be identified; this implies that molecules that undergo lipid or carbohydrate modifications in response to the environmental changes in the PV may not be detected by these methods . In addition , genetic control of immune reactivity can limit the diversity of epitopes that are generated in a given host . The later limitation can be minimized by using outbred animals ( hamsters and rabbits ) for the generation of antisera; alternatively , pooled human convalescence sera can be obtained . The expression library constructed in pET expression vectors might not be fully representative as it relies on the availability of Sau3A sites which might not be in the proximity of some genes . Lastly , as implemented here , the approach is dependent on significant difference in gene expression between axenically cultured parasites and parasites that grow within cells . There might be some benefit to using established parasite lines or parasite lines that have undergone multiple passages in axenic culture as was done in this CMAT approach . The rationale that is often stated for using parasites soon after they are recovered from the in vivo environment is that they might still be expressing in vivo induced pathogenic molecules . How soon the expression of such molecules is suppressed is however not known . If parasites that were recently established in culture were used in CMAT , the continued presence of in vivo expressed molecules might result in the removal of antibodies that are reactive to them during the adsorption protocol . Since it is not known which molecules have primary functions in the intracellular environment , it would be impossible to determine what has been lost .
|
Leishmania are intracellular parasites that reside within parasitophorous vacuoles ( PV ) in phagocytes . From within these compartments parasites control the host cell's responses to multiple stimuli . There is limited knowledge of the molecules that Leishmania parasites elaborate in the host cell to target processes therein . Furthermore , the mechanism by which such molecules would access their targets beyond the PV is not known . In the study presented here , we implemented the change mediated antigen technology ( CMAT ) to identify parasite molecules that are preferentially expressed inside infected cells . The approach was based on the reasoning that parasites express ‘new’ or antigenically modified molecules in the intracellular environment; therefore antiserum that is reactive to infected cells would contain immunoglobulins that are specific to these ‘new’ molecules . After adsorption of the antiserum with axenically cultured parasites , the antiserum was used to screen a parasite genomic expression library to identify genes encoding the molecules that are preferentially expressed in infected cells . We present for the first time evidence that some of these CMAT molecules accumulate in the PV and then traffic into the host cell in vesicles of distinct morphologies . Furthermore , several of these parasite molecules become localized in discrete compartments within the host cell .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"and",
"Discussion"
] |
[
"infectious",
"diseases/protozoal",
"infections",
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"infectious",
"diseases/skin",
"infections"
] |
2010
|
Identification of Leishmania Proteins Preferentially Released in Infected Cells Using Change Mediated Antigen Technology (CMAT)
|
A major challenge in computational neurobiology is to understand how populations of noisy , broadly-tuned neurons produce accurate goal-directed actions such as saccades . Saccades are high-velocity eye movements that have stereotyped , nonlinear kinematics; their duration increases with amplitude , while peak eye-velocity saturates for large saccades . Recent theories suggest that these characteristics reflect a deliberate strategy that optimizes a speed-accuracy tradeoff in the presence of signal-dependent noise in the neural control signals . Here we argue that the midbrain superior colliculus ( SC ) , a key sensorimotor interface that contains a topographically-organized map of saccade vectors , is in an ideal position to implement such an optimization principle . Most models attribute the nonlinear saccade kinematics to saturation in the brainstem pulse generator downstream from the SC . However , there is little data to support this assumption . We now present new neurophysiological evidence for an alternative scheme , which proposes that these properties reside in the spatial-temporal dynamics of SC activity . As predicted by this scheme , we found a remarkably systematic organization in the burst properties of saccade-related neurons along the rostral-to-caudal ( i . e . , amplitude-coding ) dimension of the SC motor map: peak firing-rates systematically decrease for cells encoding larger saccades , while burst durations and skewness increase , suggesting that this spatial gradient underlies the increase in duration and skewness of the eye velocity profiles with amplitude . We also show that all neurons in the recruited population synchronize their burst profiles , indicating that the burst-timing of each cell is determined by the planned saccade vector in which it participates , rather than by its anatomical location . Together with the observation that saccade-related SC cells indeed show signal-dependent noise , this precisely tuned organization of SC burst activity strongly supports the notion of an optimal motor-control principle embedded in the SC motor map as it fully accounts for the straight trajectories and kinematic nonlinearity of saccades .
Visually evoked saccades have remarkably stereotyped characteristics . Their two-dimensional trajectories are virtually straight , there is a near-linear relationship between movement duration and saccade amplitude , and peak eye velocity versus saccade amplitude follows a nonlinear , saturating relationship ( Fig . 1A ) . These kinematic relations , known as the ‘main sequence’ [1] , could indicate a nonlinearity in the saccadic system , because for a linear system , when driven by a step input , velocity profiles of all saccades would be self-similar , scaled ( by amplitude ) versions of each other ( see [2] for a discussion of the step-input assumption ) . As a result , the movement duration would be constant for all saccades , and peak eye velocity would increase linearly with amplitude [3] . Why saccades have these stereotyped kinematics is unknown . Interestingly , theoretical studies [4]–[8] have suggested that the main sequence of saccades could reflect an optimal control strategy , as the system has to cope with several conflicting constraints . More specifically , the properties of internal noise within the system ( assumed to increase with activity levels ) , a low spatial resolution in the peripheral retina , and a penalty for overshooting the target ( as corrective commands then have to cross hemispheres ) , require a speed-accuracy tradeoff . These studies indicated that the optimal trajectories to satisfy such constraints are met by the main-sequence relationships . However , the neural mechanisms for implementing the main-sequence relations are unknown . Almost every neural model of the saccadic system assumes that the main sequence results from a local feedback circuit in the brainstem [9]–[13] ( Fig . 1B ) . The classic theory is that this circuit receives a step input from the midbrain superior colliculus ( SC ) encoding the desired eye displacement , and that medium-lead burst cells in the pons are driven by a dynamic motor-error signal which reflects the difference between the desired and the current eye displacement . The pontine burst cells transform this signal into an eye-velocity output , a process known as pulse generation . Most saccade models assume that ( due to saturation of peak firing-rates , or neural fatigue ) the input-output characteristic of the pulse generator is a saturating nonlinearity that causes the amplitude – peak velocity relation [11]–[15] . While there is compelling evidence that the firing-rate of these neurons encodes eye velocity [11] , [16]–[19] , there is surprisingly little data to support the assumption that their input-output characteristic underlies the nonlinear main sequence . A critical problem is that the true nature and dynamics of their input signals are unknown . It is also not clear how a saturating nonlinearity in the horizontal and vertical brainstem circuits could support optimal control in two dimensions . Because the generation of straight saccades in oblique directions involves stretching the horizontal and vertical velocity components in such a way that they are scaled versions of each other ( Fig . 1A ) , an intricate cross-coupling between the horizontal and vertical pulse generators would be required [20]–[25] . Clearly , straight trajectories are optimal in the sense that they constitute the shortest path to the endpoint . Here , we study the role of the superior colliculus in the optimal control of saccades . The deeper layers of the SC form a topographic map of saccade vectors ( Figs . 1B and 1C ) , which is organized in eye-centered coordinates [26] . Neurons in this motor map fire a brisk burst of action potentials tightly coupled to the onset and duration of the saccade . Cells near the rostral pole of the SC are involved in the generation of small saccades , while cells at caudal sites encode large eye movements . Saccade direction in the contralateral hemifield is represented along the medial-lateral dimension . The range of movement vectors for which an SC neuron is recruited is called its movement field [27] , [28] , and it has been inferred from the size and shape of the cells' movement fields that each saccade is associated with a location-invariant two-dimensional Gaussian activation profile in the map [28] . However , how the temporal dynamics of activity within the active population contributes to the movement kinematics is not clear [29]–[33] . Because the SC is an important sensorimotor interface in the control of saccade behavior [34]–[36] , it could be in an ideal position to optimize speed-accuracy tradeoff . In line with this notion , we recently found that the nonlinear kinematics of saccades are already embedded in the spatial-temporal SC activity patterns [33] . More specifically , we demonstrated that measured SC firing patterns , when used to drive a linear feedback model of the brainstem , produce realistic , straight saccades that obey the nonlinear main sequence . As shown in Fig . 1C , this analysis assumed that i ) the spatial-temporal activity patterns in the SC motor map are decoded by a linear ‘spike-vector’ summation mechanism in which each spike from each neuron adds an independent , site-specific vectorial contribution to the saccade command , and that ii ) this command is executed by two independent , linear feedback circuits in the brainstem that control the horizontal and vertical movement components , respectively . Thus , none of the nonlinear properties of saccades ( component stretching , skewness of velocity profiles , and main-sequence relations ) were built into the model; they all emerged from the measured SC activity patterns ( see also Video S1 ) . However , the mechanism by which the recruited SC population generates these properties remained unclear . In a recent theoretical study we proposed that a possible mechanism could reside in a topographic organization of burst properties of saccade-related SC cells [37] . This theory predicts that the SC could specify the nonlinear main sequence of saccades if the saccade-related bursts vary from brief and intense in the rostral zone , to less intense and of longer duration in the caudal zone , while keeping the total number of spikes constant . It also predicts a systematic rostral-to-caudal increase in burst skewness , because skewness of the eye-velocity profiles increases with saccade amplitude [38] . Moreover , if the SC indeed acts as an optimal controller , one would expect that cells in the recruited population synchronize their burst profiles because this would ensure that the net SC movement command specifies a straight trajectory at optimal speed . Note , however , that if burst durations and skewness indeed vary along the map , such synchronization of burst activity can only occur if the shape of each cell's burst depends on the planned movement , rather than on its anatomical location in the motor map . We thus predicted that burst shapes are different depending on whether a cell is recruited for small versus large saccades . Finally , optimal control theories predict that speed-accuracy tradeoff is constrained by noise , which increases with the amplitude of the control signals [5]–[7] . This suggests that SC cells might exhibit signal-dependent noise . Several studies have examined the spatial-temporal organization of saccade-related burst activity in the SC [29]–[33] , but a detailed , quantitative analysis is still missing . Furthermore , the possibility that populations of movement cells might encode the optimal kinematics of different movements through a topographic gradient of firing properties has never been studied . In the present paper we therefore characterized the activity patterns of a large population of saccade-related SC neurons , widely distributed across the motor map . Our results reveal a highly systematic organization of burst properties along the rostral-to-caudal extent of the SC motor map , which can fully account for the nonlinear kinematics of saccades , their straight trajectories in oblique directions , and the skewed shape of their velocity profiles . Moreover , we demonstrate signal-dependent noise in the number of spikes of saccade-related bursts , as predicted by optimal-control theories . These remarkable findings strongly support the notion of an optimal motor-control principle embedded in the SC .
One of the central premises in the optimal control theories proposed by Harris and Wolpert [5] and Tanaka et al . [7] is that the amount of noise in the control signal is proportional to signal amplitude , but experimental evidence for this assumption is still limited [39]–[41] . Here we test whether such signal-dependent noise exists at the level of the SC motor map . Towards that end , we quantified the trial-to-trial variability of saccade-related burst activity of SC neurons for visually guided saccades into their movement field . Figure 2 illustrates the results for a typical cell . The left-hand panels show the instantaneous firing rate of the cell ( color code ) as a function of time for selected saccades of different amplitudes ( Fig . 2A , ‘amplitude scan’ ) and directions ( Fig . 2B , ‘direction scan’ ) . Individual trials are sorted according to saccade amplitude and direction , respectively . For each saccade , the burst magnitude was quantified from the raw data by counting all spikes in the time window between the two tick marks ( identifying saccade onset an offset with a 20 ms lead time ) , and the resulting spike counts are displayed in the adjacent panels as running averages . Saccade endpoints of the responses in Figs . 2A and B ( squares and circles , respectively ) are plotted in Fig . 2C , together with a two-dimensional representation of the cell's movement field ( Methods ) . This movement-field plot shows that the number of spikes in the burst ( color code ) varies systematically with saccade amplitude and direction . However , as can be seen in Figs . 2A and B , it is not only the mean number of spikes in the burst ( open symbols ) that depends systematically on the amplitude and direction of the saccade vector; also the trial-to-trial variability in the spike counts changes systematically ( error bars indicate ±1 SD ) . More specifically , when the spike-count variability is plotted as a function of the average number of spikes in the burst ( Fig . 2D ) , it appears for this cell that the standard deviation increases almost linearly with the mean . To quantify this relation , we fitted a regression line to the data . Note that the intercept is practically zero . The slope of the regression line thus provides a good measure of the so-called coefficient of variation , which is the ratio between the standard deviation and the mean ( Methods ) . For the neuron in Fig . 2 , the coefficient of variation was Cv = 0 . 278±0 . 016 ( mean±SEM ) . To quantify the signal-dependent variability of saccade-related bursts for the entire population of cells , we selected for each SC neuron a series of non-overlapping clusters of closely matched saccade responses . This analysis , which is illustrated in Fig . 3A and B , typically included the entire movement field . Figures 3C and D summarize the results for all 108 cells for which we obtained sufficient data ( see Methods , for details ) . Note that for the vast majority of cells ( 89% , n = 96 ) spike-count variability increased significantly with the mean number of spikes ( two-tailed t-test , p<0 . 05 ) . The mean ( ±SEM ) value of the coefficient of variation across the cell population was 0 . 306±0 . 014 . These findings thus indicate that motor commands generated by individual SC neurons are endowed with a considerable amount of signal-dependent noise . Because signal-dependent noise alone cannot fully account for the variability in observed saccade trajectories , it has been suggested that cell activity may also be endowed with signal-independent noise [42] . We therefore quantified the level of signal-independent noise in the SC by analyzing the intercepts of the regression lines ( c . f . , Fig . 2D and Fig . 3B ) . Averaged across our sample of cells , the mean intercept was indeed significantly different from zero ( two-tailed t-test , p<0 . 001 ) , but with an average value of 0 . 54±0 . 03 spikes ( mean±SEM ) it reached only 3 . 0±0 . 2% of the cells' average peak response . As mentioned in the introduction , we recently proposed on the basis of model simulations that a topographic organization of burst properties within the SC motor map could underlie the nonlinear kinematics of saccades [37] . More specifically , our simulations showed that the SC population could specify the main sequence if the saccade-related bursts vary from brief and intense in the rostral zone ( small-saccade area ) , to less intense and of longer duration in the caudal zone ( encoding large saccades ) while keeping the number of spikes constant . To test these theoretical predictions , Fig . 4 quantifies several burst properties of saccade-related SC cells for saccades towards the center of their movement field as a function their anatomical rostral-to-caudal location in the motor map: number of spikes in the burst ( Fig . 4A ) , mean firing rate ( Fig . 4B ) , and peak firing rate ( Fig . 4C ) . Spike counts and mean firing rates were calculated from the raw data , while peak firing rates were estimated from the spike-density functions . We indexed the rostral-to-caudal location of each cell in the motor map by the amplitude of its preferred vector . Cells ( n = 103 ) were selected for having at least 5 saccades into the center of their movement field ( Methods ) . Note that the spike counts are remarkably constant across the rostral-to-caudal extent of the SC as cells at each location fire on average about 18 spikes for their preferred saccade . Mean and peak firing rates , on the other hand , decrease systematically with the rostral-to-caudal location of the cells . To quantify these topographic relations we fitted linear regression lines ( solid ) to each of the three datasets . The numerical results are listed in Table 1 ( top entries ) . This regression analysis showed there is no significant correlation between spike counts and preferred amplitudes , and thus anatomical location , of the neurons ( Pearson correlation , p>0 . 3 ) . Correlations between firing rates and preferred amplitudes , however , were significantly different from zero ( t-test , p<0 . 005 ) . Slopes of the regression lines ( mean±SEM ) in Fig . 4B and C were −7 . 6±1 . 9 and −7 . 3±2 . 3 spk/s per deg , respectively . The data thus indicate an almost two-fold decrease in firing rates from small to large saccades within the oculomotor range . Because mean firing-rates were computed from the number of spikes divided by saccade duration ( Methods ) , a reduction in mean firing-rate with amplitude could , in principle , result from the confounding factor that saccade duration increases linearly with amplitude ( c . f . , Fig . 1A ) . However , this potential confound does not play a role for the peak firing-rate . In our analysis , the absolute peak firing-rate values are partly determined by the width of the Gaussian smoothing kernel ( see Table 1 for results with different values of ) , but for a fixed kernel width ( here , = 5 ms ) , the changes in peak firing-rate across the population of cells are entirely due to changes in the temporal distribution of spikes within their bursts . The data in Fig . 4C thus confirm the presence of a real rostral-to-caudal gradient in the cells' activity patterns , which – to our knowledge – has not been documented before . This spatial gradient is further illustrated in Figs . 5A–D which show the full temporal burst profiles of four clusters of cells that fired ∼18 spikes for their preferred saccade ( dashed circles in Fig . 4A ) . Insets illustrate , schematically , the location of the recording sites in the motor map . It is important to note that this analysis does not involve any normalization of activity . Thin lines represent the mean spike-density functions of the individual cells , where data are aligned with saccade onset and averaged across at least 5 saccades into the center of the movement field ( within 0 . 5⋅σmf ) . Amplitudes of the preferred vectors were about 5° , 13° , 21° , and 32° , respectively , as may be inferred from the corresponding eye-position traces . Thick lines are the grand averages of the responses . Note that the systematic rostral-to-caudal changes in mean and peak firing rates are quite obvious from these discharge profiles too . Cells recorded in the rostral region of the SC ( Fig . 5A ) clearly showed much higher peak firing rates and shorter bursts for their preferred vector than the ones in the caudal SC ( Fig . 5D ) , while cells found at intermediate locations ( Fig . 5B and 5C ) had intermediate firing rates and burst durations . In Fig . 6A we normalized the average burst profiles from the four clusters of cells ( Figs . 5A–D ) with respect to their peak . The top-right insets show the main sequence behavior and velocity profile skewness [38] of the corresponding average eye movements ( data replotted from Fig . 5 ) . This analysis shows that the peak firing rate occurs at about the same instant relative to saccade onset ( see Table 1 , for further quantification ) while the burst duration increases systematically with the preferred amplitude of the cells ( indexed by the gray-scaling ) . Hence , these results indicate that , just like the saccade velocity profiles , the skewness of the bursts increases systematically with saccade amplitude ( and duration ) . Note , however , that the shapes of the spike-density waveforms in Fig . 6A also depend on the width of the Gaussian smoothing kernel that is used to compute the instantaneous firing rates ( here , = 5 ms ) . To circumvent this problem , in Fig . 6B , we therefore calculated burst skewness directly from the distribution of spike moments in the burst ( Methods ) , rather than from the spike density functions . As in Fig . 4 , cells ( n = 103 ) were selected for having at least five saccades into the center of their movement field . Note that bursts associated with small saccades ( produced by cells in the rostral region of the SC ) are nearly symmetric ( skewness about zero ) while bursts associated with large saccades ( produced by cells in the caudal SC ) have longer tails towards the end of the saccade ( positive skewness ) . Linear regression analysis showed a significant , positive correlation ( t-test , p≪0 . 001 ) between the skewness of the bursts and the preferred amplitudes of the cells . This shows that burst skewness indeed increases systematically with the cells' rostral-to-caudal location in the motor map . Table 1 summarizes the numerical results of this regression analysis ( fifth entry ) . The systematic increase in burst skewness with saccade amplitude was not related to any change in spike-count variability; the coefficient of variation computed for these bursts ( Eq . 4 ) was more or less constant across the map ( Table 1 ) , and did not significantly correlate with burst skewness ( t-test , p>0 . 28 ) . The topographic variations in burst properties that are revealed by the analyses in Figs . 4–6 were consistent across animals . In all three animals for which we obtained a sufficient number of cells at different locations in the map ( n>10; smallest eccentricity <4°; largest eccentricity >40° ) , we observed a nearly constant number of spikes across the map , a systematic rostral-to-caudal decrease in mean and peak firing rates , and a systematic rostral-to-caudal increase in burst skewness . In the fourth animal , recording sites did not span a large enough range of eccentricities ( 11°<R<15° ) to obtain reliable fit results . So far the analyses quantified the burst properties of SC neurons for saccades towards the center of their movement field , i . e . , when the cells are part of the central region of the recruited population . However , SC cells are recruited for a wide range of saccades ( the movement field ) that all have different kinematics depending on their amplitude . The question therefore arises what happens to the burst of a given cell when it is recruited for saccades that deviate from its preferred vector . To address this question , we examined the shape of the burst profiles for saccades towards more eccentric locations within the movement field . More specifically , we compared the burst dynamics when the same cells are part of the rostral region of the recruited population ( i . e . they are recruited for saccades that are larger than their preferred movement ) , with the burst dynamics when they belong to the caudal region of the recruited population ( for saccades that are smaller than the preferred one; see motor-map insets Figs . 7 and 8 , for schematic illustration ) . If saccade-related SC cells indeed participate in specifying the optimal saccade trajectory [33] , one would expect that the shape of their bursts is different depending on whether the cells are recruited for saccades that are smaller than their preferred vector , versus whether they are recruited for saccades that are larger than their preferred vector , because the shapes of the velocity profiles for smaller and larger saccades differ ( Fig . 6 , inset ) . Alternatively , the shape of the bursts could remain the same regardless of the actual saccade vector . Such behavior is expected if the burst activity of each cell represents only the weight of its preferred vector in a downstream center of gravity computation of the saccade goal [43]–[45] . In this case , the number of spikes produced by each cell is determined by its input , but the temporal firing properties only depend on its location in the motor map . To dissociate between these two possibilities , Figs . 7 and 8 show the measured burst profiles of two cells in for a series of large and small saccades for which the cell fired the same number of spikes . We chose two cells from the central region of the SC ( preferred amplitudes 9 . 5° and 13 . 5° , respectively ) because these cells fired vigorous bursts for saccades with distinctly different kinematics . Note that the shapes of the bursts for small saccades ( light-gray traces ) and large saccades ( dark-gray traces ) are clearly different . For both cells , peak firing rates occurred at about the same instant relative to movement onset regardless of the movement amplitude , but burst durations were shorter and peak firing rates were higher when the cell took part in the small saccades than when it participated in the large saccades . These examples thus suggest that the shape parameters of the burst depend systematically on the actual saccade to which the neuron contributes , rather than on its topographic location in the motor map . If the SC acts as an optimal controller , one would expect that cells in the recruited population synchronize their burst profiles because this would ensure that the net SC movement command specifies a straight trajectory with optimal horizontal and vertical velocity components . The example cells in Figs . 7 and 8 suggest that the burst shape of individual SC cells is indeed tuned such that all cells in the recruited population become synchronized for the saccade in which they participate . To determine whether burst synchronization in the recruited population is actually found across the SC , we studied the spatial-temporal burst dynamics of all recorded cells in the motor map for a range of different saccade amplitudes ( Fig . 9 ) and directions ( Fig . 10 ) , cross-correlated the burst shapes of all recruited cells to determine their similarity ( Fig . 11 ) , and analyzed the burst shapes of a fixed set of neighbouring neurons for different saccade vectors ( Fig . 12 ) . We first examined the temporal discharge profiles of saccade-related neurons along the rostral-to-caudal extent of the SC for saccades of five particular amplitudes . Because we applied the rotation algorithm ( Methods ) , cells having different preferred directions could all be pooled . Figure 9 shows this analysis for saccade amplitudes ranging from 5 degrees ( top panels ) to 31 degrees ( bottom panels ) . Saccade directions always corresponded with the direction of the cells' preferred movement ( Methods ) . The different plots in Fig . 9 thus provide estimates of the rostral-to-caudal cross-sections through the center of the population . As in previous figures , we indexed the rostral-to-caudal location of each cell by the amplitude of its preferred vector . The color-codes in Fig . 9A reflect the averaged spike density of individual cells as a function of time relative to saccade onset and their location in the SC motor map . The associated mean eye movements are superimposed . We calculated the activity profiles in Fig . 9B from the raw data in Fig . 9A by averaging the spike-density functions of nearby cells according to a Gaussian weighting function ( width σ = 0 . 25 mm ) and sorting them according to their location in the motor map ( Methods ) . Bin centers were chosen at 2 . 5 deg intervals in visual space . In Fig . 9C , the site-dependent spike density functions are collapse onto a single pair of axes , and bin centers were chosen at 0 . 2 mm intervals on the SC motor map . The latter produced spike density functions of the population activity at equally spaced distances from the center of the active population . To facilitate visual inspection , the hue of the individual spike density functions indicates the rostral ( green ) to caudal ( cyan ) location of the cells while the color saturation is proportional to the firing rate ( i . e . , traces become darker when activity increases ) . Note that for a given saccade amplitude the burst profiles along the rostral-to-caudal extent of the SC appear to have very similar shapes , which change systematically as function of saccade amplitude ( and duration ) . More specifically , peak firing rates in the recruited population decrease systematically with increasing saccade amplitude while burst duration and skewness increase with increasing saccade amplitude . The data thus demonstrate that not only the location of the recruited population changes systematically with saccade amplitude; also the dynamics of the burst activity within the active population changes systematically . The observation that cells at different locations within the active population all have very similar burst dynamics is further illustrated in Fig . 10 . As in Fig . 9 , we plot the responses of cells as a function of their rostral-to-caudal location in the SC , but now for 14 degree amplitude saccades in five different directions , ψ , relative to their preferred vector . The plots in Fig . 10 thus provide rostral-to-caudal cross sections through the population for a 14 degree saccade at five different iso-direction lines . Iso-direction line ψ = 0 deg runs through the center of the active population , so the plots in the center row of Fig . 10 are equivalent to the ones in the center row of Fig . 9 . The other iso-direction lines , however , characterize the temporal discharge profiles at different medial-lateral locations . Note that the peak firing rates decrease systematically with increasing rostral-to-caudal and increasing medial-to-lateral distance from the center of the population activity while the shape of the temporal burst profiles remains remarkably similar . In fact , it is not difficult to see that practically all discharge profiles in Fig . 10 are approximately scaled versions of each other . Our findings thus demonstrate that the temporal dynamics of burst activity is very similar throughout the population of recruited cells . To quantify the temporal synchrony and shape similarity of the burst profiles at different locations within the active population , we performed a series of temporal cross-correlation analyses ( see Methods , for details ) . As shown in Fig . 11A , we first normalized the site-dependent spike density functions from Fig . 9C with respect to their peak . For each saccade amplitude , we then cross-correlated the population activity at the center of the activated region of cells with the population activity at different rostral-to-caudal distances from the center ( solid lines ) , and with the activity of the individual cells ( open symbols ) . Note that for each of the five movement amplitudes the normalized population responses fall on top of each other ( Fig . 11A ) . Accordingly , the cross-correlation analyses performed on the population data ( solid lines ) produced correlation values at lag zero that were close to one ( Fig . 11B ) , and optimal delays that were close to zero ( Fig . 11C ) . These results thus indicate that the response profiles are indeed synchronized , scaled versions of each other . Even at the level of the individual cells it is observed that the cross-correlation values at lag zero are very high ( gray squares; typically r ( τ = 0 ) >0 . 8 ) . For about 50–70% of the cells , the cross-correlation values at the optimal delay ( black circles ) were significantly higher than at lag zero ( Fig . 11B ) , but the burst delays of the individual cells were not systematically related to their rostral-to-caudal location within the recruited population ( Fig . 11C ) . The same pattern of results was obtained along the medial-lateral dimension of the population ( data not shown ) . The cross-correlation values themselves were of course influenced to some extent by the width of the Gaussian smoothing kernel ( here , = 5 ms; Methods ) , but the resulting optimal delays were not . When we repeated the analysis with different kernel widths ( in the range of 2 to 10 ms ) , they were virtually identical . The robust analysis of relative burst delays in Fig . 11 thus demonstrates that there is no systematic spread of activity , neither in the rostral-to-caudal direction [31] , [46] ( which would produce a systematic increase of the lag ) , nor from the center towards the periphery of the population [27] , [47] ( which would produce a V-shape pattern ) . Note , however , that the response profiles for saccades of different amplitudes are clearly different . I . e . , burst profiles within the active population become more and more skewed as saccade amplitude and duration increases . The interesting question then arises whether these systematic changes in burst shape are also reflected in the population activity at a given location in the SC when the cells at that location participate in the generation of saccades of different amplitudes ( and durations ) . The results of Figs . 7 and 8 , in which we analyzed the responses of two example cells for small versus large saccades , would indeed suggest such changes . To address this question , we selected in Fig . 12 a cluster of 16 neurons located in the central region of the SC . Preferred amplitudes of these cells were closely matched , ranging between 13–15 degrees . From the responses of the individual cells ( color code in Fig . 12A ) we first calculated the mean population response at that location for five different saccade amplitudes ( solid lines in Fig . 12A ) , and we then normalized the resulting response profiles with respect to their peaks ( Fig . 12C ) . Note that there are systematic increases in burst duration and skewness as the saccade amplitude increases from 8 to 32 degrees . Also note the main-sequence behavior in the corresponding eye movement traces of Fig . 12B .
We previously showed that a simple linear ensemble-coding model of the SC motor map could fully account for the nonlinear properties of saccades [33] . By driving a linear model of the brainstem with actual recorded spike trains from a large population of SC neurons widely distributed across the motor map we obtained realistic , straight saccades with the correct kinematic properties ( Video S1 ) . The assumptions of this linear ensemble-coding model ( Fig . 1C ) are the following: Note that our scheme does not make any prior assumptions about the activity patterns of individual cells in the SC motor map . For example , the finding that the number of spikes in the burst is invariant for fast and extremely slow eye movements [54] , and that it is invariant across the motor map ( Fig . 4A ) are not properties of the model , but appear to be properties of the motor map . The same holds for our current findings that the peak firing rate of cells , their burst duration and their burst skewness all vary in a systematic way with their location in the motor map . Finally , that all cells within the population are synchronized during the saccade , and that the burst properties are determined by the saccade in which the cell participates , rather than by its location in the map , is not a model assumption either . Yet , all these features taken together fully explain how the SC population could encode the nonlinear kinematics , and at the same time generate straight saccades in all directions . The nonlinear behavior of saccade kinematics is due to two opposing factors: ( i ) peak firing rates in the caudal SC are lower than in the rostral SC , but ( ii ) each spike in the caudal SC has a much larger impact on the brainstem than a rostral cell , due to the exponentially growing efferent mapping function . Thus , although spike firing-rates decrease in the caudal SC , the increase in eye displacement provided by each spike is much larger at these locations . Straight saccades result from synchronization of burst profiles , especially along the medial-lateral dimension of the SC ( Fig . 10 ) . This synchronization ensures that the horizontal and vertical velocity commands are scaled versions of each other as is required for producing straight oblique saccades . Our findings thus strongly support the idea that the SC motor map acts as a nonlinear vectorial pulse generator . Because the saccade results from the linear contribution of a large ensemble of recruited cells , our linear ensemble-coding model does not necessarily predict that each individual cell should encode the details of the saccade kinematics . Nevertheless , the saccade kinematics are to a large extent reflected at the level of single cells , as the shape of the saccade-related burst follows a similar skewness relationship with burst duration as the saccade velocity profile to saccade duration ( Fig . 6 ) . In addition , the burst shape of any individual cell is to a large extent determined by the saccade metrics , and can thus vary substantially between small and large saccades into its movement field ( Figs . 7 , 8 and 12 ) . Interestingly , these features are predicted by the notion that saccade-related SC neurons have dynamic movement fields which determine the dynamic relationship between the activity of individual cells and the ensuing eye displacement as a function of the saccade metrics [33] . This concept also predicts that for movements to a single visual target the recruited cells act together as a ‘common source’ by synchronizing their burst profiles . This behavior is indeed observed ( Figs . 9–11 ) . Taken together , our findings provide strong support for the argument that the nonlinear saccade kinematics are not due to a passive saturation of brainstem burst neurons , e . g . , as a result of neural fatigue , but reflect a deliberate design property within the saccadic system to produce the main-sequence characteristics . We believe that such a strategy aims to optimally cope ( in a statistical sense ) with conflicting constraints that impede the fovea from getting as fast and as accurately as possible on a peripheral target of interest . Several constraints may be identified within the system: neural noise , considerable processing delays , and the highly inhomogenous organization of the retina that introduces considerable uncertainty about stimulus locations within the visual periphery . Indeed , theoretical studies on optimization have provided an elegant explanation for different features of saccadic behavior , such as the tendency to systematically undershoot visual targets [55] , but also the main-sequence nonlinearity [4]–[7] . In these studies , the noise is assumed to be multiplicative , i . e . signal-dependent . To optimize such a system for speed and accuracy , the control signal should obey the nonlinear main sequence , and at the same time employ an undershooting strategy . Our data show for the first time that SC cells indeed possess multiplicative , signal-dependent noise ( Figs . 2 and 3 ) , and that this property is invariant across the motor map ( Table 1 ) . Since the coefficient of variation is , on average , constant across the map , it follows from our spike-vector summation theory that the variance in the resulting displacement vector , , as function of desired amplitude and direction is given by:Where Ni ( R , Ф ) is the mean number of spikes fired by cell i , its spike-vector contribution , and Cv the coefficient of variation ( population average ) . Clearly , the covariances in this equation cannot be determined from our single-unit recordings , so we cannot be 100% sure that the net collicular output has multiplicative , signal-dependent noise . Simulations with a uniform motor map ( i . e . , fixed cell density , , and location-invariant widths σmf and heights Npref of the Gaussian movement fields ) showed , however , that both correlated and uncorrelated cell activities produce elliptical endpoint distributions ( e . g . , [56] ) with standard deviations that increase linearly with eccentricity ( not shown ) . The optimal control models described above all dealt with the generation of horizontal saccades . However , in two dimensions the fastest response should also follow a straight line; thus an extended optimal control theory would predict straight oblique saccades [8] . In alternative models , in which the brainstem is driven by separate horizontal and vertical nonlinear pulse generators , the programming of straight oblique saccades is highly nontrivial as it requires a tedious scheme of cross-coupling between the horizontal and vertical systems [21]–[23] . The problem is even more complex when considering head-free gaze shifts [57]–[59] , because a fixed cross-coupling scheme will no longer work when the eye-head coupling varies considerably from trial to trial . However , if the nonlinear burst generator resides in the SC motor map , straight saccades ( and head-free gaze shifts ) become an emerging property of the system , requiring no further cross-coupling than a simple linear horizontal/vertical decomposition ( i . e . sine and cosine projection ) of the vectorial gaze-shift command . Our simple linear ensemble-coding model is still incomplete . It accounts for the generation of saccadic eye movements once the spatial-temporal distribution of SC activity is known , but it does not explain how the tuned burst patterns come about in the first place ( an omission in most models of the saccadic system ) . In principle , the saccade-related burst of SC neurons could be derived from two alternative mechanisms . First , the type of input signal , such as that from the frontal eye fields ( FEF ) , might impose the spatial-temporal pattern of excitation [60] . However , this explanation is neither attractive nor plausible; it merely shifts the problem to a different area of the brain , and there exists ( at least to our knowledge ) no evidence that the FEF is involved in the dynamic , online control of saccade trajectories under normal conditions . Second , the burst might arise from intrinsic membrane properties of saccade-related SC neurons [61] , or from properties of the local circuits within the SC motor map . Recent in vitro experiments indeed suggest that the synchronous bursting command observed in our data could result from a local excitatory network , in combination with NMDA receptor activation in the deeper layers of the SC [62]–[64] . These in vitro experiments also demonstrated a strong nonlinear signal amplification process in the SC , which is interesting because it might account for the nonlinearity ( i . e . , vector averaging ) of responses obtained with certain types of electrical double stimulation [65]–[67] . In theory , feedback from the brainstem saccade generator could also contribute to the shaping of the burst dynamics . For example , Van Opstal and Kappen [68] suggested that a linear model of the brainstem together with weighted feedback projections to the SC motor map reproduces straight saccades with the correct kinematics . However , in their scheme caudal cells receive the weakest feedback , and the model therefore predicts a strongly asymmetric distribution of burst durations and skewness within the recruited population . This is clearly not observed in our data ( Figs . 9–11 ) . Moreover , previous perturbation studies have shown that activity in the SC is also not consistent with other types of feedback models , as the SC activity does not encode dynamic motor error [32] , [54] , [69]–[74] . Further research is needed to elucidate the mechanisms that shape the spatial-temporal firing patterns during saccades , and the behavior of the SC population in more complex motor behaviors , like during head-free gaze shifts , curved double-step saccades , or electrical microstimulation . Nevertheless , the burst properties reported in this study strongly support the idea that the deeper layers of the SC act as an optimal controller: the systematic organization of peak firing rates and burst durations as function of saccade amplitude along the motor map , the synchronous change in firing rate of recruited cells in the population , and the shaping of the temporal burst profile of a given cell with the currently planned saccade , all contribute to the generation of straight eye-movement trajectories with optimal kinematics .
We collected data from four rhesus monkeys ( macaca mulatta ) that were trained to follow a small visual target with saccadic eye movements . The setup , surgical procedures and behavioural paradigms have been described elsewhere [33] , [54] , [71] . In short , the animals were seated in a primate chair facing a spherical array of light-emitting diodes ( LEDs ) in an otherwise completely dark room . The head was restrained , and movements of the eye were measured with the double-magnetic induction technique [75] , [76] . Single-cell recordings were made through a recording cylinder using tungsten microelectrodes that were advanced into the SC with the use of a hydraulic stepping motor . We recorded activity of 146 saccade-related neurons that were found in the intermediate and deep layers of the SC ( about 0 . 5–3 mm below the dorsal surface ) . Cells were considered saccade-related if they showed an increase in firing rate around the onset of saccades towards a particular region of the visual field . All of these cells were studied with a standard saccade task in which the animal made saccades from an initial fixation LED to a peripheral target LED which was presented for 500 ms . The movement field of each neuron was determined by eliciting saccades to targets inside and neighboring the response field ( ‘movement field scan’ ) . In addition , saccades were evoked to a fixed series of targets across the visual field ( R between 2–35 deg , Φ∈[0 , 30 , … , 360] deg; ‘rose scan’ ) . Eye movement data , spike data , and movement field parameters of all 146 recording sites were stored in a database for further processing . The file contained data from a total of 32 , 147 trials . Saccades were detected off-line on the basis of the calibrated eye-position signals using custom software ( see [71] for details ) . Subsequent analysis was done in Matlab 7 . 9 ( version R2009b ) . Single-cell activity was displayed in spike rasters and peri-stimulus time histograms ( PSTHs ) that were aligned on specific events such as target onset and the onset of a saccade . Spike trains from individual trials were represented as a sequence of δ pulses at the time of spike occurrence , ( 1 ms resolution ) : . Spike density functions , , were calculated from these raw spike trains by convolving with a Gaussian smoothing kernel that had a default width of = 5 ms and a height of spk/s , but analyses were also repeated for different kernel widths ( see text ) .
|
As the fovea is the only spot on the retina with high spatial resolution , primates need to move their eyes to peripheral targets for detailed inspection . Saccades are the fastest movements of the body , and theoretical studies suggest that their trajectories are optimized to bring the fovea as fast and accurately as possible on target . Speed-accuracy optimization principles explain the stereotyped nonlinear ‘main-sequence’ relationship between saccade amplitude , duration , and peak velocity . Earlier models attributed these kinematic properties to nonlinear neural circuitry in the brainstem but this creates problems for oblique saccades . Here , we demonstrate how the brainstem can be linear , and how instead the midbrain superior colliculus ( SC ) could optimize saccadic speed-accuracy tradeoff . Each saccade involves the recruitment of a large population of SC neurons . We show that peak firing-rate and burst shape of the recruited cells systematically vary with their location in the SC , and that burst shapes nicely match the eye-velocity profiles . This organization of burst properties fully explains the main-sequence . Moreover , all cells synchronize their bursts , thus maximizing the total instantaneous input to the brainstem , and ensuring that oblique saccades have straight trajectories . We thus discovered a sophisticated neural mechanism underlying optimal motor control in the brain .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"anatomy",
"and",
"physiology",
"electrophysiology",
"neuroscience",
"motor",
"systems",
"cognitive",
"neuroscience",
"computational",
"neuroscience",
"neurological",
"system",
"coding",
"mechanisms",
"robotics",
"musculoskeletal",
"system",
"biology",
"systems",
"biology",
"visual",
"system",
"physiology",
"sensory",
"systems",
"computational",
"biology",
"motor",
"reactions",
"neurophysiology",
"genetics",
"and",
"genomics"
] |
2012
|
Optimal Control of Saccades by Spatial-Temporal Activity Patterns in the Monkey Superior Colliculus
|
Cell-surface transferrin receptor ( CD71+ ) erythroid cells are abundant in newborns with immunomodulatory properties . Here , we show that neonatal CD71+ erythroid cells express significant levels of V-domain Immunoglobulin ( Ig ) Suppressor of T Cell Activation ( VISTA ) and , via constitutive production of transforming growth factor ( TGF ) - β , play a pivotal role in promotion of naïve CD4+ T cells into regulatory T cells ( Tregs ) . Interestingly , we discovered that CD71+VISTA+ erythroid cells produce significantly higher levels of TGF-β compared to CD71+VISTA− erythroid cells and CD71+ erythroid cells from the VISTA knock-out ( KO ) mice . As a result , CD71+VISTA+ erythroid cells—compared to CD71+VISTA− and CD71+ erythroid cells from the VISTA KO mice—significantly exceed promotion of naïve CD4+ T cells into induced Tregs ( iTreg ) via TGF-β in vitro . However , depletion of CD71+ erythroid cells had no significant effects on the frequency of Tregs in vivo . Surprisingly , we observed that the remaining and/or newly generated CD71+ erythroid cells following anti-CD71 antibody administration exhibit a different gene expression profile , evidenced by the up-regulation of VISTA , TGF-β1 , TGF-β2 , and program death ligand-1 ( PDL-1 ) , which may account as a compensatory mechanism for the maintenance of Treg population . We also observed that iTreg development by CD71+ erythroid cells is mediated through the inhibition of key signaling molecules phosphorylated protein kinase B ( phospho-Akt ) and phosphorylated mechanistic target of rapamycin ( phospho-mTOR ) . Finally , we found that elimination of Tregs using forkhead box P3 ( FOXP3 ) -diptheria toxin receptor ( DTR ) mice resulted in a significant expansion in the frequency of CD71+ erythroid cells in vivo . Collectively , these studies provide a novel , to our knowledge , insight into the cross-talk between CD71+ erythroid cells and Tregs in newborns . Our results highlight the biological role of CD71+ erythroid cells in the neonatal period and possibly beyond .
Reticulocytes originate in bone marrow ( BM ) from erythroblasts by the process of nuclear extrusion and are released into the blood , where they further mature into erythrocytes ( red blood cells ) . Reticulocytes contain RNA , mitochondria , endoplasmic reticulum , and ribosomes . As reticulocyte maturation progresses , the nucleus vanishes , protein synthesis ceases , and cell-surface transferrin receptor ( CD71 ) disappears [1] . The main function of vertebrate erythrocytes has been considered to be oxygen transporters; however , other functions such as interactions with immune cells and immunomodulatory properties have also been attributed to their immature counterparts [2 , 3] . In 1979 , for the first time , Pavia and colleagues demonstrated the immunosuppressive effects of murine neonatal splenocytes on adult cells [4 , 5] . Subsequent studies towards understanding the immunomodulatory mechanism of these cells showed that TGF-β [6] and direct cell–cell interactions might partially be involved for their immunosuppressive effects [7] . Moreover , it has been suggested that erythroid precursors promote polarization of naive CD4+ T cells to different T cell subtypes [3 , 8] and may alter the T-helper 1/2 ( Th1/Th2 ) cytokine balance toward a Th2 phenotype in newborns [9] . In agreement , we have recently discovered that CD71+ erythroid cells are enriched in newborns and , because of their immunosuppressive properties , impair both the innate and adaptive immune responses in newborns [10–12] . These cells coexpress CD71 ( transferrin receptor ) and glycophorin-A–associated protein ( TER119; the erythroid lineage marker ) in mice or CD71 and CD235a ( the erythroid lineage marker ) in humans [11] . In addition , we have shown that CD71+ erythroid cells express arginase-2 , and this enzymatic activity is required for their immunosuppressive activity [10 , 11] . Overall , our previous studies have revealed that neonatal vulnerability to infection results from the temporary existence of immunosuppressive CD71+ erythroid cells . In agreement , a recent study has shown a marked expansion of erythroid precursors ( CD71+ TER119+ ) in murine spleens following Salmonella infection and demonstrated that presence of these cells was associated with enhanced bacterial persistence [9] . More recently , the immunomodulatory properties of CD71+ erythroid cells in cord blood of preterm versus full-term newborns have been investigated [13] . Although we have shown that CD71+ erythroid cells express arginase-2 , it is unclear whether these erythroid cells express surface molecules such as inhibitory receptors in order to execute their immunosuppressive/immunoregulatory properties . Inhibitory receptors or immune checkpoints regulate T cell function by restraining initial T cell activation that otherwise can be potentially pathogenic for the host . However , sustained expression of coinhibitory receptors such as program death-1 ( PD-1 ) , cytotoxic T-lymphocyte-associated protein 4 ( CTLA-4 ) , T-cell immunoglobulin and mucin-domain containing-3 ( TIM-3 ) , and lymphocyte activation 3 ( LAG-3 ) are hallmarks of dysfunctional T cells [14–16] . Coinhibitory receptors , upon interaction with their ligands on antigen-presenting cells ( APCs ) and/or tumor cells , mediate immune suppression [14 , 15] . V-domain Ig Suppressor of T Cell Activation ( VISTA ) , a newly discovered inhibitory receptor , is a transmembrane immunoglobulin ( Ig ) super family also known as Programmed Death-1 Homologue-1 ( PD-1H ) , vsir , Gi24 , Dies1 , DD1α , c10orf54 , and SISP1 [17–19] . VISTA messenger RNA ( mRNA ) is expressed predominantly in hematopoietic and less in nonhematopoietic tissues . Within the hematopoietic compartment , VISTA is expressed on dendritic cells , monocytes , neutrophils , and natural killer ( NK ) cells , as well as naïve T cells and regulatory T cells ( Tregs ) , but not on B cells [17–19] . So far , several studies have shown that VISTA has immunomodulatory functions , regardless of its regulatory role in differentiation of osteoblast , adipocyte , and embryonic stem cells [20–23] and cell apoptosis [24] . VISTA is unique among other immunoglobulin superfamily molecules but lacks classic immunoreceptor tyrosine-based inhibition motifs ( ITIMs ) or immunoreceptor tyrosine-based switch motifs ( ITSMs ) [18] . This explains the dual role for VISTA on APCs or T cells as a ligand or receptor , respectively . The inhibitory function of VISTA has widely been studied and supported by several in vitro and in vivo reports [18 , 19 , 25 , 26] . These studies have revealed that agonistic monoclonal antibodies to VISTA induce inhibitory response [26] , in contrast to antagonistic antibodies that promote stimulatory action [18 , 25 , 27] . Although VISTA is considered as an inhibitory molecule , no evidence of overt autoimmunity was reported in VISTA knock-out ( KO ) mice . However , T-cell–activation phenotype and accelerated aging is reported in these VISTA KO mice [26 , 28 , 29] . Furthermore , the anti-inflammatory role of VISTA has been proved in different studies using mouse models of Graft Versus Host Disease ( GVHD ) , Concanavalin A ( Con-A ) –induced hepatitis [28 , 29] , and experimental autoimmune encephalomyelitis [27] . Moreover , it has been shown that VISTA has an anti-inflammatory role via regulation of the interleukin-23/17 ( IL-23/IL-17 ) axis [30] , and VISTA blockade in various animal tumor models has shown promising results in terms of enhanced T cell response against the tumor [25 , 26 , 31 , 32] . However , the mechanism ( s ) underlying the role of VISTA in stem cell differentiation and/or immunomodulation have yet to be well elucidated . Some reports indicated that VISTA induces stem cell differentiation via interaction with TGF-β family members including Bone Morphogenic Protein-4 ( BMP4 ) . In the TGF-β downstream pathway , VISTA promotes BMP4 signaling through activation of SMAD 1 , 5 , and 8; VISTA forms a complex with SMAD 4 and regulates target gene expression toward stem cell differentiation [20 , 22 , 33] . Moreover , VISTA with microRNA-125a ( miR-125a ) generates a regulatory loop that modulates BMP4 signaling on stem cell differentiation in vitro [22] . Other studies suggest that VISTA is constitutively expressed on Tregs and plays an important role in their function [20 , 22] . As such , VISTA improves the induction of Foxp3+-induced Tregs ( iTregs ) in the presence of TGF-β in both mice and human CD4+ T cells in vitro [18 , 32] . A recent study using wild-type ( WT ) and VISTA KO mice has shown that VISTA is required for de novo induction and expansion of iTregs from naïve T cells [34] . Since some reports indicated a role for VISTA in stem cell differentiation , we aimed to investigate the expression and function of VISTA in CD71+ erythroid cells . In this study , for the very first time to our knowledge , we have shown that CD71+ erythroid cells express significant levels of VISTA , and more importantly , the CD71+VISTA+ subpopulation , compared to the CD71+VISTA− subpopulation , produces significant levels of TGF-β . In addition , we have shown defects in CD71+ erythroid cells from VISTA KO mice in terms of TGF-β production ability . We also demonstrate that CD71+VISTA+ erythroid cells via TGF-β enhance iTreg induction in vitro . Thus , our studies provide a novel insight , to our knowledge , into the role of CD71+ erythroid cells in immune regulation and cross-talk with Tregs .
In order to determine the transcriptional profile of CD71+ erythroid cells over time , we conducted RNA sequencing ( RNAseq ) analysis on total RNA extracted from enriched CD71+ erythroid cells from the spleen of 3- , 6- , 12- , and 28-day-old BALB/c mice . When hierarchical clustering was conducted on Euclidian distances between samples , day 3 CD71+ erythroid cells clearly showed a different gene expression profile than the rest of the time points ( Fig 1A ) . For the most part , day 6 and day 12 samples formed separated branches on a dendrogram , while day 28 samples exhibited a more erratic distribution along such dendrograms ( Fig 1A ) . Those results are partially recapitulated in principal component analysis ( PCA ) on the Euclidian distances between samples . In essence , day 3 CD71+ erythroid cells clearly separated from the rest of the samples along the first principal component , and day 6 samples formed a rather discrete cluster . Day 12 and day 28 samples intermingled ( Fig 1B ) . In summary , the transcriptional profile of day 3 samples is clearly distinct from the rest of the samples . Day 6 CD71+ erythroid cells were found to have a transcriptional profile somewhat different from day 12 and day 28 samples , and the latter two groups showed a similar transcriptional profile . This suggests that a rather drastic change in the gene expression program of CD71+ erythroid cells takes place early during mice development , and no other major changes are observed at later time points . We have previously shown that CD71+ erythroid cells are enriched in C57BL/6 newborn mice and gradually disappear by 3 weeks of age [11] . To determine the frequency of CD71+ erythroid cells in the spleens of three different mice strains ( BALB/c , C57BL/6 , and filial 1 hybrid [F1] ) , we compared their presence at different ages ( S1A Fig ) . The percentages of CD71+ erythroid cells in BALB/c peaked between days 6 and 9 , followed by a gradual decline at later time points to reach levels comparable to adult mice by day 28 ( S1A and S1B Fig ) . A similar pattern was observed for the frequency of CD71+ erythroid cells in C57BL/6 and F1 mice , respectively ( S1C and S1D Fig ) . Of note , although the frequency of CD71+ erythroid cells reached adult levels by day 21 in Cincinnati [11] , these cells remained for a week longer in the neonatal mice in Edmonton . This might be due to differences in microbiome or other physiological and environmental factors . We decided to determine gene expression of different inhibitory molecules and/or their ligands in CD71+ erythroid cells . Our RNAseq data indicated high gene expression for galectin-9 ( Lgals9 ) , galectin-1 ( Lgals1 ) , and VISTA ( vsir ) in CD71+ erythroid cells , respectively ( Fig 1C ) . However , other inhibitory receptors were absent or expressed at very low levels ( e . g . , galectin-3 [Lgals3] , TIGIT , LAG-3 , and PD-L1 ) ( Fig 1C ) . To confirm our RNAseq data , we measured surface expression of highly expressed inhibitory molecules ( Lgals9 , Lgals1 , and VISTA ) on CD71+ erythroid cells . We found that VISTA was highly expressed on CD71+ erythroid cells compared to Lgals9 and Lgals1 ( Fig 1D–1F ) . Therefore , we aimed to investigate the role of VISTA in CD71+ erythroid cell function . As shown in Fig 2A and 2B , approximately 40% of CD71+ erythroid cells expressed VISTA in neonatal BALB/c mice at day 3 . VISTA expression levels peaked on days 6–9 ( >60% ) , followed by a significant decline at day 12 , which remained at similar levels thereafter . In contrast , in the C57BL/6 mouse strain , there was a lower expression level for VISTA on CD71+ erythroid cells in 3-day-old mice ( <30% compared to >40% in BALB/c mice ) and significantly increased in 6-day-old mice , still lower than the same age in BALB/c mice ( approximately 45% compared to approximately 65% ) but peaked to a similar level as BALB/c mice at day 9 , followed by a significant decline at day 12 ( Fig 2C ) . Contrary to BALB/c mice , VISTA expression was significantly increased at days 15–18 in C57BL/6 mice and then remained constant until day 28 ( Fig 2C ) . In F1 mice , approximately 50% of CD71+ erythroid cells expressed VISTA at day 3 , and the expression level of VISTA peaked at day 6 but , interestingly , was reduced at day 9 compared to BALB/c and C57BL/6 mice ( Fig 2D ) . The expression level of VISTA then decreased abruptly to reach the minimum level at day 12 , after which there was a significant increase at day 15 , which remained at a similar level until day 28 ( Fig 2D ) . In addition , the intensity of VISTA expression on CD71+ erythroid cells was measured at different time points . We observed that the mean fluorescence intensity ( MFI ) of VISTA was significantly increased on CD71+ erythroid cells of neonatal mice at days 9 and 15 compared to day 3 in both BALB/c ( Fig 2E and 2F ) and B57BL/6 mice ( Fig 2G and 2H ) . However , the intensity of VISTA on CD71+ erythroid cells in F1 mice remained unchanged at days 3 , 9 , and 15 ( Fig 2I and 2J ) . In order to confirm the RNAseq data , we evaluated VISTA gene expression levels in mice at different ages . Consistent with the protein levels , VISTA mRNA expression levels were significantly higher at days 6 and 9 compared to day 3 , followed by a significant drop at day 12 in BALB/c mice ( Fig 2K ) . We further analyzed the expression of VISTA among CD71+ erythroid cells obtained from the BM compared to the spleen . Interestingly , we found that the MFI of VISTA on CD71+ erythroid cells from BM was significantly higher than their counterparts in the spleen ( Fig 2L and 2M ) . Similar results were obtained for the percentages of VISTA+CD71+ erythroid cells in the BM versus the spleen ( S1E Fig ) . Although adult mice have very low percentages of CD71+ erythroid cells in their spleens compared to neonates ( S1F Fig ) , these cells expressed high levels of VISTA , but no difference in its expression level was observed when male and female mice were compared ( S1G Fig ) . What is the role of VISTA on CD71+ erythroid cells ? Given the association of VISTA with TGF-β family members , including BMP4 , we investigated TGF-β–associated gene expression profile in CD71+ erythroid cells . Interestingly , we found TGF-β1 , Smad3 , Smad5 , and TGF-β receptor 1 ( r1 ) genes were markedly increased ( Fig 3A ) . Increased TGF-β expression by CD71+ erythroid cells was also confirmed by quantitative polymerase chain reaction ( qPCR ) and flow cytometry ( Fig 3B and 3C ) . We then investigated whether there was any association between VISTA expression and TGF-β production . We found that CD71+VISTA+ cells constitutively produce significantly higher levels of TGF-β compared to CD71+VISTA− cells at different ages in BALB/c mice ( Fig 3D and 3E ) . Similar results were observed among neonatal CD71+ erythroid cells in C57BL/6 and F1 mice ( Fig 3F and 3G , respectively ) . To determine whether we can use Toll-like receptor agonists to enhance TGF-β secretion from CD71+ erythroid cells , we analyzed the expression of different toll-like receptors on these cells . We found that CD71+ erythroid cells express Toll-like receptor ( TLR ) 1 , TLR2 , and TLR-4 genes . However , their expression levels dropped significantly at day 3 compared to day 1 , and by day 6 , their expression levels became almost undetectable and remained at very low levels thereafter ( Fig 3H , 3I and 3J ) . Therefore , we decided to measure the expression of TLRs among neonatal CD71+ erythroid cells at day 3 by flow cytometry . As shown in Fig 3K and 3L , we observed significantly higher surface expression of TLR1 among neonatal CD71+ erythroid cells in liver and spleen compared to their counterparts in the BM and thymus . Since CD71+ erythroid cells express TLR1 , we then stimulated CD71+ erythroid cells with N-palmitoyl-S-dipalmitoylglyceryl ( Pam3 ) Cys-Ser- ( Lys ) 4 ( CSK4 ) , a TLR-1 agonist , and heat-killed ( HK ) Listeria monocytogenes ( Lm ) , a TLR-2 agonist . As is shown in Fig 3M , stimulation by HK Lm and/or Pam3CSK4 enhanced production of TGF-β among CD71+ erythroid cells . Of note , since we have shown a small portion of CD71+ erythroid cells express Lgals1 and Lgals9 on their surface ( Fig 1E and 1F ) , we compared TGF-β production by Lgals9+ versus Lgals9− cells . We found that Lgals9+ cells produce higher levels of TGF-β compared to Lgals9− CD71+ erythroid cells . However , VISTA+CD71+ erythroid cells surpass Lgals9+ cells in terms of TGF-β production ( Fig 3N ) . In addition , we observed coexpression of VISTA with Lgals1 and Lgals9 on a small portion of CD71+ erythroid cells ( S1H and S1I Fig ) . Furthermore , we investigated the expression of glycoprotein A repetitions predominant ( GARP ) on VISTA+ versus VISTA− CD71+ erythroid cells . Although VISTA+ cells appear to express slightly higher percentages of GARP , it did not reach significance ( S1J and S1K Fig ) . We first determined the frequency of Tregs at different ages in newborn BALB/c mice . Tregs can be detected in the spleen of the newborn mouse starting at day 3 . The percentage of Tregs rose significantly thereafter until day 9 , after which there was a gradual increase to reach the adult levels ( Fig 4A and 4B ) . However , the absolute number of Tregs expanded gradually from the day 3 to day 18 and then significantly increased at days 21 and 28 before reaching the adult level ( S1L Fig ) . As shown in Fig 4B , we observed a dramatic increase in the percentages of Tregs at days 6 and 9 , which coincided with the maximal abundance of CD71+ erythroid cells in newborn mice ( S1A–S1D Fig ) . Based on these observations , we decided to determine whether there was cross-talk between these two immunosuppressive cell populations in the neonatal period . As we have shown in Fig 3D , VISTA+CD71+ compared to VISTA−CD71+ erythroid cells produce significantly higher levels of TGF-β . Therefore , we isolated VISTA+CD71+ and VISTA−CD71+ erythroid cells and cocultured each subpopulation with the isolated naïve CD4+ T cells from adult mice at a 1:1 ratio . After 4 days , we found that VISTA+CD71+ erythroid cells promoted naïve CD4+ T cell conversion into iTregs ( Fig 4C and 4D ) . Our data demonstrate that VISTA+CD71+ erythroid cells significantly expanded Tregs mainly via TGF-β production because blocking TGF-β significantly abrogated the effects of VISTA+CD71+ on iTreg induction ( Fig 4C and 4D ) . In contrast , the effects of VISTA−CD71+ subpopulation on the induction of Tregs was significantly lower ( Fig 4C and 4D ) , possibly due to the lower TGF-β production by these cells . Next , the phenotype of iTregs was analyzed , and we found iTregs do not express TIGIT and express low CD73 but high PDL-1 , VISTA , Lgals9 , and CD39 ( Fig 4E ) . Interestingly , we found both natural and iTregs express helios ( Fig 4F ) , which agrees with the recent finding claiming that helios should not be considered as a marker of natural Tregs [35 , 36] . Furthermore , we found that these iTregs significantly inhibited the proliferative capacity of effector T cells in a dose-dependent manner when compared to the spleen Tregs ( Fig 4G and 4H ) . Other studies have shown that antagonizing the Akt signaling pathway is required for Treg induction [37 , 38] . To test this , we cultured naïve T cells in the presence or absence of CD71+ erythroid cells for 18 h and then measured phosphorylation of Akt and mTOR . Intracellular staining for phospho-Akt and phospho-mTOR revealed significant reduction in Akt ( Fig 4I and 4J ) and mTOR phosphorylation ( Fig 4K and 4L ) levels when naïve T cells were cocultured with CD71+ erythroid cells . In addition , we decided to investigate whether CD71+ erythroid cells impact FOXP3 induction via arginase-2 expression in naïve CD4+ T cells in vitro . We have reported elsewhere that CD71+ erythroid cells utilize this enzyme as one of their immunosuppressive mechanisms [11] . Therefore , we measured FOXP3 expression on naïve CD4+ T cells in the presence of CD71+ erythroid cells and L-arginine . Our observations showed that neutralization of arginase-2 by L-arginine supplementation did not impact Treg frequency ( S1M Fig ) . To better understand the function of VISTA , we investigated the role of CD71+ erythroid cells from VISTA KO mice for the induction of Tregs . We found no significant difference in the frequency of CD71+ erythroid cells between WT and VISTA KO mice at ages of 6 and 9 days ( Fig 5A and 5B ) . However , we found impaired TGF-β production ability of CD71+ erythroid cells from VISTA KO compared to the WT mice ( Fig 5C and 5D ) . Since we have shown in Fig 4C and 4D that CD71+VISTA+ erythroid cells , compared to CD71+VISTA− erythroid cells , were superior in Treg induction , we analyzed the effects of CD71+ erythroid cells from VISTA KO mice compared to WT mice on FOXP3 induction in vitro . Interestingly , we found that CD71+ erythroid cells from VISTA KO were defective in FOXP3 induction when cocultured with naïve CD4+ T cells in vitro ( Fig 5E and 5F ) . Consistently , lower Treg frequency was observed in VISTA KO compared to the WT mice ( Fig 5G and 5H ) . Finally , the expression of GARP on CD71+ erythroid cells in VISTA KO and WT mice was measured . Interestingly , CD71+ erythroid cells from VISTA KO mice had significantly higher levels of GARP compared to the WT group ( Fig 5I and 5J ) . Of note , an abundance of activated immune cells , CD71+TER119− , in the spleen of VISTA KO mice compared to the WT group was evident ( Fig 5K and 5L ) . CD71 ( transferrin receptor ) is an activation marker and can be expressed on different immune cells including T cells [39] . It appears that these activated nonerythroid lineage cells in VISTA KO mice consist of different immune cells , but the majority were CD3+ T cells ( Fig 5K ) . To further establish the relationship between CD71+ erythroid cells and Tregs , CD71+ erythroid cells were depleted using anti-CD71 antibody . As we have previously described , this antibody depletes 50%–60% of CD71+ erythroid cells in vivo [10–12] . Since CD71+ erythroid cells reach to the maximal levels at days 6–9 in newborns , we decided to deplete these cells by a single injection of anti-CD71 antibody ( approximately 150 μg ) at day 9 , and 2 days later , we investigated the frequency of Tregs and CD71+ erythroid cells in their spleens . As shown in Fig 6A and 6B , administration of anti-CD71 antibody reduced approximately 60% of CD71+ erythroid cells in the spleen of mice . However , depletion of CD71+ erythroid cells had no significant effects on the percentage of Tregs ( Fig 6C and 6D ) , which is in contrast to our in vitro observation ( Fig 4C and 4D ) . Although Treg frequency was unchanged , potential changes in Treg phenotype following CD71+ erythroid cell depletion was studied . We observed that Tregs exhibited significantly higher MFI for CD25 and Ki67 in the absence of CD71+ erythroid cells , respectively ( S2A and S2B Fig ) . In addition , significant up-regulation of PDL-1 and GARP on Tregs following anti-CD71 treatment was noted ( S2C–S2F Fig ) . However , no changes in TIGIT or CTLA-4 expression levels in Tregs under these circumstances were observed ( S2G–S2J Fig ) . These phenotypical changes in Tregs suggest a possible feedback when CD71+ erythroid cells were depleted . Since depletion of CD71+ erythroid cells did not impact Treg frequency in vivo , we decided to better characterize the remaining and/or newly generated CD71+ erythroid cells following anti-CD71 antibody administration by conducting RNAseq analysis . Interestingly , we observed that the remaining and/or newly generated CD71+ erythroid cells have a different transcriptional profile than their counterparts in the control animals ( Fig 6E ) . Anti-CD71–treated animals significantly up-regulated VISTA expression compared to the IgG isotype control group in both the percentages and MFI of VISTA expression ( Fig 6F–6H ) . In addition , CD71+ erythroid cell depletion led to a rise in the gene expression of both TGF-β1 and TGF-β2 by the remaining or newly produced CD71+ erythroid cells ( Fig 6E and 6I and 6J ) . More importantly , we observed that CD71+ erythroid cells from anti-CD71-antibody–versus rat immunoglobulin G- ( IgG ) isotype–treated group significantly up-regulated TLR2 but not TLR4 mRNA expression levels ( Fig 6K and 6L ) . In agreement , we found higher TLR2 surface expression on these cells compared with their native counterparts ( Fig 6M and 6N ) . Based on these data , we decided to determine whether remaining and/or newly generated CD71+ erythroid cells respond to HL Lm , a TLR2 agonist , and subsequently produce more TGF-β compared to controls . Interestingly , we found this was the case , and CD71+ erythroid cells from the anti-CD71–treated group produced significantly higher TGF-β when stimulated with HK Lm in vitro ( Fig 6O and 6P ) . Furthermore , we found up-regulation of inhibitors of DNA binding ( ID ) differentiation genes ( Id1 and Id2 ) in CD71+ erythroid cells obtained from the anti-CD71–treated mice compared to the control group , which are downstream of the BMP4–VISTA pathway ( Fig 6E , 6Q and 6R ) . Thus , higher VISTA expression may result in up-regulation of Id1 and Id2 genes , which is in line with a report indicating down-regulation of both Id1 and Id2 mRNA levels when the VISTA gene is silenced [20] . In addition , up-regulation of PDL-1 , galectin-3 ( Lgals3 ) , and Lgals1 in CD71+ erythroid cells from the treated animals with anti-CD71 antibody versus controls was observed ( Fig 6E and 6S–6U ) . Despite the fact that depletion of CD71+ erythroid cells did not impact Treg frequency in vivo , we decided to determine what happens to the frequency of CD71+ erythroid cells when Tregs are depleted in newborns . Therefore , frequency of CD71+ erythroid cells was analyzed in the presence and absence of Tregs using FOXP3-DTR B57BL/6 mice . FOXP3-DTR mouse allows elimination of FOXP3+ following administration of diphtheria toxin ( DT ) ( 35 ng/g body weight ) [40] . For these studies , we decided to use 11-day-old neonatal mice that had substantial percentages of Tregs . Two consecutive injections of DT at days 11 and 12 , as anticipated , resulted in elimination of the majority of Tregs ( Fig 7A ) . Subsequently , we found that elimination of FOXP3+ Tregs led to a significant increase in the percentages of CD71+ erythroid cells in these mice ( Fig 7B and 7C ) . Finally , in order to exclude possible adverse effects of DT on the erythropoiesis in neonates , we administered DT ( 35 ng/g body weight ) into C57BL/6 WT mice and measured the frequency of CD71+ erythroid cells following two consecutive treatments . As shown in Fig 7D and 7E , administration of DT did not change the frequency of CD71+ erythroid cells in the WT neonatal mice . To further understand the mechanisms whereby depletion of Tregs led to expansion of CD71+ erythroid cells , we measured VISTA expression on CD71+ erythroid cells . Interestingly , we did not observe any significant difference in VISTA expression levels between control and Treg-depleted groups ( Fig 7F and 7G ) . In addition , we cocultured isolated CD71+ erythroid cells and Tregs to determine possible effects of Tregs on CD71+ erythroid cells in vitro . Interestingly , we observed that addition of Tregs to CD71+ erythroid cells resulted in a significant decrease in MFI of CD71 ( transferrin receptor ) on these erythroid cells ( Fig 7H and 7I ) . In addition , presence of Tregs down-regulated the MFI of Ki67 on CD71+ erythroid cells in vitro ( Fig 7J and 7K ) . However , no changes in the expression of other inhibitory receptors ( e . g . , VISTA and PDL-1 ) were observed ( S2K–S2N Fig ) . These observations suggest possible cross-talk between these two cell populations during the neonatal period . We , and others , have shown that CD71+ erythroid cells are abundant in human cord blood and placenta tissues and coexpress CD71 and CD235α [11 , 41–43] . Thus , we decided to determine whether our observations were reproduceable in human samples . We found that human cord blood CD71+ erythroid cells express much lower levels of VISTA compared to their counterparts in mice . Interestingly , expression of VISTA was significantly higher on placenta CD71+ erythroid cells compared to the cord blood ( Fig 7L and 7M ) . In addition , we performed qPCR for the expression of VISTA ( PD-1H ) gene in enriched cord blood and placenta CD71+ erythroid cells . Interestingly , we were able to detect PD-1H gene in both cord blood and placenta CD71+ erythroid cells . However , its expression level was significantly higher in placenta CD71+ erythroid cells compared to the cord blood ( Fig 7N ) . Since human CD71+ erythroid cells get lysed when exposed to the perm buffer for intracellular cytokine staining , we performed qPCR on enriched CD71+ erythroid cells for TGF-β gene expression . We observed that both cord blood and placenta CD71+ erythroid cells express TGF- β; however , gene expression of this cytokine was significantly higher in placenta than cord blood CD71+ erythroid cells ( Fig 7O ) .
In this report , we demonstrate a novel role , to our knowledge , for CD71+ erythroid cells in promoting the development of iTregs . Consistent with our previous studies [10 , 11] , we have shown that CD71+ erythroid cells are physiologically enriched in newborns and gradually disappear by day 28 . The transcriptional analysis confirms a different transcriptional profile for CD71+ erythroid cells in the neonatal period . This suggests that a rather drastic change in the gene expression program of CD71+ erythroid cells takes place early during development . However , transcriptome analysis determines that these cells consistently express some functional genes associated with inhibitory receptors and/or their ligands such as Lgals9 , Lgals1 , and VISTA . In agreement , we find VISTA is highly expressed on neonatal CD71+ erythroid cells despite some fluctuations in its expression levels throughout the first 4 weeks of life . Although Lgals9 and Lgals1 genes are highly expressed in these cells , surface expression of these molecules is substantially low on CD71+ erythroid cells compared to the VISTA . Here , we demonstrate that neonatal VISTA+CD71+ erythroid cells are responsible for promoting iTreg development . In contrast , VISTA−CD71+ erythroid cells and CD71+ erythroid cells from VISTA KO mice show significant defect in promotion of naïve CD4+ T cells into FOXP3+ iTregs when cocultured in vitro in the presence of low-dose IL-2 . This observation confirms a role for VISTA in the conversion of naïve CD4+ T cells into iTregs . The roles of different inhibitory molecules in the development and maintenance of Tregs have been widely described . For instance , CTLA-4 is required for TGF-β–mediated induction of FOXP3+ Tregs from CD4+CD25− cells [44] , and Lgals9 enhances iTreg stability and function via interaction with CD44 , which forms a complex with TGF-β r1 [45] . Likewise , PDL-1 promotes the induction and maintenance of iTregs [37] . Although PDL-1 signaling alone is sufficient to promote iTreg development and can promote iTreg development in the absence of TGF-β [37] , this is not the case for VISTA [34] . Recently , it has been reported that compared to naïve T cells from WT mice , stimulation of naïve T cells from VISTA KO mice with anti-CD3/CD28 in the presence of TGF-β resulted in decreased induction of iTregs [34] . The induction of iTregs was nearly abolished in both WT and VISTA KO naïve T cells in the absence of TGF-β . However , the suppressive function of iTregs was not affected by the loss of VISTA [34] . Our transcriptome analysis determines that CD71+ erythroid cells consistently express TGF-β , which was confirmed by qPCR and flow cytometry analysis . More importantly , CD71+VISTA+ erythroid cells are the dominant TGF-β–producing cells compared to their VISTA− counterparts or CD71+ erythroid cells from VISTA KO mice . In agreement , we find that CD71+VISTA+ erythroid cells significantly promote FOXP3 expression in naïve CD4+ T cells , demonstrating their TGF-β–dependent effects on the development of iTregs . This finding is further reinforced by the observation that inhibition of TGF-β abrogates FOXP3 expression in naïve CD4+ T cells when cocultured with CD71+VISTA+ erythroid cells . These observations demonstrate that VISTA is associated with the TGF-β overproduction phenotype in CD71+ erythroid cells . It is worth noting that Lgals9+CD71+ erythroid cells also contribute to TGF-β production despite their low frequency . Thus , our studies reveal a novel mechanism , to our knowledge , by which CD71+VISTA+ erythroid cells mediate immune tolerance . We have already demonstrated that CD71+ erythroid cells have immunosuppressive properties in the neonate [10–12] and play an essential role in fetomaternal tolerance [41 , 46] . CD71+ erythroid cells get expanded in the peripheral blood of women during the course of pregnancy; however , this was not the case when we compared the frequency of these cells in healthy women versus women with inflammatory bowel disease ( IBD ) [46] . Interestingly , we found reduced frequency of CD71+ erythroid cells in the peripheral blood of IBD women was associated with a reduction in Tregs in these patients [46] . Taken together , here we provide a novel role , to our knowledge , for CD71+ erythroid cells in immune hemostasis by promoting iTregs from naïve CD4+ T cells . However , CD71+ erythroid cells in human cord blood and/or placenta express much lower VISTA on their surface compared to their counterparts in neonatal mice . Where do CD71+VISTA+ erythroid cells exert their crucial role on iTreg development ? CD71+ erythroid cells are mostly abundant in spleen , BM , and blood but in much lower frequency in lymph nodes ( approximately 5% ) of newborns [10–12] . On the other hand , VISTA is widely expressed on hematopoietic and to a lesser extent on nonhematopoietic cells [19] , and overexpression of VISTA is associated with a reduction in T cell activation and proliferation and with reduced cytokine production [18 , 19] . In addition , Wang and colleagues have shown that PD-1HIg ( VISTA ) promotes induction of iTregs in the presence of TGF-β [34] , which is in agreement with our finding that CD71+VISTA+ erythroid cells , via TGF-β , promote development of iTregs from naïve CD4+ T cells . In addition , this group indicated that this effect , largely mediated via suppression of inflammatory cytokines as such VISTA signaling , prevents the conversion of iTregs to Th1 and Th17 in an inflammatory condition [34] . Although this is not investigated in our current study , we have previously reported the inhibitory effects of CD71+ erythroid cells on the production of Th1 and Th17 cytokines in an infection model [12] . We , and others , have shown that CD71+ erythroid cells from mice or human cord blood suppress cytokine production and T cell proliferation in vitro [10 , 11 , 13] and utilize arginase-2 as one of their potential immunosuppression mechanisms [10 , 11] . However , it should be noted that the role of CD71+ erythroid cells in suppression of immune responses may be more complex and possibly would work through multiple mechanisms . Although arginase-1 , via dendritic cells , can promote FOXP3 induction [47] , our observations do not support a role for arginase-2 in Treg development . Intriguingly , we reveal a crucial role for CD71+VISTA+ erythroid cells in the control of iTreg pool size in newborns , which may contribute to the inhibition of T cell responses . In agreement , lower Treg population and subsequently abundance of activated immune cells , especially T cells , was noted in VISTA KO mice . Although the percentages of VISTA-expressing cells among CD71+ erythroid cell population vary from day 3 to 28 , consistently their frequency peaks at days 6–9 in different mice strains , which coincides with the similar pattern , expansion , in Treg frequency in mice . This may suggest cross-talk between CD71+ erythroid cells and Tregs in the newborn . We see a dramatic reduction in the expression of VISTA on CD71+ erythroid cells in all three mice strains at age 12 . The postnatal development of the hypothalamic-pituitary-adrenal ( HPA ) axis in mice occurs at two different stages . Day 12 marks the transition time when the visual stimulatory signals occur . After the stress-hyporesponsive period ( day 12 ) , mice exhibit enhanced corticosterone basal levels and a response of adrenocorticotropic hormone ( ACTH ) and corticosterone [48] . Some changes , such as up-regulation of VISTA , may occur following the visual stimulatory signals after day 12 , when mice open their eyes . In agreement , it has been shown that corticosteroid treatment leads to the up-regulation of coinhibitory molecules such as CTLA‐4 , PD‐1 , CD73 , and FOXP3 in a colitis model [49] . Although it is plausible that the increase in corticosteroids after day 12 enhances the expression of VISTA on CD71+ erythroid cells , the interaction of corticosteroids and VISTA merits further investigation . VISTA is a crucial element of BMP4 signaling , suggesting that it acts as a BMP4 coreceptor [20] . In fact , VISTA has the highest expression levels in the BM and spleen [19] , where immune cells including CD71+ erythroid cells are present . In our complementary studies , we decided to find out whether depletion of CD71+ erythroid cells impairs Treg development in vivo . Although complete depletion of CD71+ erythroid cells , because of their nature ( being red blood cell precursors ) , is impossible , their partial depletion did not impact Treg frequency in vivo . This suggests the existence of a differential mechanism in place by the remaining and/or newly generated CD71+ erythroid cells to compensate for the loss of their depleted siblings . In agreement , we find that the remaining and/or newly generated CD71+ erythroid cells express significantly higher levels of VISTA and subsequently elevated expression levels of both TGF-β1 and TGF-β2 genes . Smads 2 and 3 are receptor-regulated Smads and promote TGF-β signal via interactions with Smad 4 [50] . In contrast , Smads 6 and 7 are hindrance Smads and act to repress the TGF-β signal by competing with receptor-regulated Smads for the receptor ( TβR ) [51] . TGF-β overproduction can increase growth promotion by suppressing Id genes , which is the case in normal cells [50] . Conversely , in some malignant cells , unlike primary cells , Ids are not down-regulated by TGF-β [52] . Similarly , our results indicate up-regulation of Id1 and Id2 genes despite a spike in TGF-β production in CD71+ erythroid cells following anti-CD71 administration , suggesting that Ids can function either in concert with or opposition to TGF-β function . Although no significant difference in the Smad genes was observed in CD71+ erythroid cells from control versus anti-CD71–treated mice , PDL-1 was highly up-regulated in newly generated and/or remaining CD71+ erythroid cells post treatment . This may explain an alternative compensatory mechanism for CD71+PDL-1+ erythroid cells to promote iTreg development because PDL-1 plays a critical role in the development and maintenance of iTreg function , especially mediating immune regulation where TGF-β is present [37] . Although neonatal CD71+ erythroid cells do not express PDL-1 , we have shown that pregnancy-induced CD71+ erythroid cells ( either in spleen or placenta ) express substantial levels of PDL-1 and/or PDL-2 [41] . Furthermore , expression of TLRs on CD71+ erythroid cells and changes in their expression levels following anti-CD71 treatment reveal a novel role , to our knowledge , for CD71+ erythroid cells in sensing pathogen-associated molecular patterns ( PAMPs ) . The abundance of CD71+ erythroid cells in the newborn and their ability to recognize PAMPs illustrate a dynamic role for these cells in the neonatal period . We also find that CD71+ erythroid cells attenuate the Akt signaling pathway during the conversion of naïve CD4+ T cells to iTregs by reducing the phosphorylation of Akt and its downstream substrate mTOR . In agreement , previous studies have shown that truncation of TCR signaling and inhibition of Akt and mTOR signaling axis are crucial for the development of iTregs [38 , 53 , 54] . This negative regulation is mediated by the neuropilin-1-semaphorin-4a axis for phosphatidylinositol 3-kinase-AKT ( PI3K-Akt ) [55] and of the mTOR complex 2 ( mTORC2 ) pathway by the inositol phosphatase phosphate and tensin homolog ( PTEN ) [56] in Tregs . Despite the fact that depletion of CD71+ erythroid cells did not influence Treg frequency in newborn mice , it potentially impacts their proliferative capacity , as shown by enhanced CD25 and Ki67 expression in Tregs in the absence of CD71+ erythroid cells . In addition , we decided to determine the effects of Treg depletion on the frequency of CD71+ erythroid cells in vivo . We find depletion of Tregs using FOXP-3-DTR mice results in a significant increase in the percentages of CD71+ erythroid cells . Although depletion of Tregs does not impact VISTA expression on CD71+ erythroid cells in vivo , we see significant reduction in the expression of CD71 and Ki67 in CD71+ erythroid cells when cocultured with Tregs in vitro . These observations demonstrate cross-talk between these two immunosuppressor cell populations . While revealing the molecular mechanism of this observation merits further investigations , we suggest depletion of Tregs may impact erythropoiesis . For instance , IL-2 KO mice that have lower Treg frequency experience anemia [57] . This might be due to the immune activation in the absence of Tregs and subsequently extramedullary erythropoiesis , which results in the expansion of immature red blood cells in the periphery . These data suggest that both CD71+ erythroid cells and Tregs , by utilizing different mechanisms , may contribute to the immune regulation in the newborn . Thus , our study highlights the important role of CD71+ erythroid cells in the neonatal period .
All animal experiments were carried out in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the Canadian Council for Animal Care . The experimental protocol was approved by the Committee on the Ethics of Animal Experiments at the University of Alberta ( Protocol # AUP00001021 ) . Male and female BALB/c and C57BL/6 mice were purchased from the Charles River Institute ( Morrisville , NC , USA ) . BALB/c and C57BL/6 mice were bred together to create F1 mice . In addition , B6 . 129 ( Cg-FOXP3 DTR ) mice were purchased from The Jackson Laboratory ( Bar Harbor , ME , USA ) . All animals were maintained and bred under pathogen-free conditions within the animal care facility at the university of Alberta . Similarly , C57BL/6 VISTA KO mice were maintained and bred under pathogen-free conditions within the animal care facility at the VA Medical Centre , White River Junction , VT , USA . For in vivo depletion , purified anti-CD71 ( 8D3 ) and rat IgG2a isotype control antibodies were administered i . p . ( 150–200 µg ) at day 6 , and 2 days later , spleens were harvested for immunological assays . For Treg depletion , FOXP3-DTR neonatal mice ( 9 days old ) were injected i . p . with DT ( Sigma-Aldrich ) , 35 ng/g body weight , before sample collection at day 11 . Fluorophore or biotin-conjugated antibodies with specificity to mouse cell antigens and cytokines were purchased from eBioscience ( Waltham , MA , USA ) or BD Biosciences ( San Jose , CA , USA ) . Specifically , the following antibodies were used for mouse studies: anti-CD71 ( C2 ) , anti-VISTA ( MIH64 , MIH63 ) , anti-LAP ( TW7-16B4 ) , anti-CD3 ( 500A2 ) , anti-CD4 ( RM4-5 ) , anti-CD62L ( MEL-14 ) , anti-CD44 ( 5035–41 ) , anti-TER-119 ( TER-119 ) , anti-CD73 ( Ty/23 ) , anti-CD39 ( 24DMS1 ) , anti-PDL-1 ( MIH5 ) , anti-TIGIT ( 1G9 ) , anti-Helios ( 22F6 ) , anti-CD25 ( PC61 . 5 ) , anti-FOXP3 ( FJK-165 ) , GARP ( YGIC86 ) , IgG1 K ISO control ( P3 . 6 . 2 . 8 . 1 ) , anti-CD281 ( TLR1 , eBioTR23 ) , and anti-CD284 ( TLR4 , MT5510 ) . Recombinant mouse TGF-β1 , anti-Lgals9 ( RG-395 ) , and anti-CD282 ( TLR2 , T2 . 5 ) were purchased from Biolegend ( San Diego , CA , USA ) , and TGF-β1 blocker was purchased from Sigma-Aldrich ( St . Louis , MO , USA ) . Anti-Lgals1 ( EPR3205 ) and the secondary antibody ( A31572 ) were purchased from Abcam ( Cambridge , UK ) . Human antibodies against CD71 ( MA712 ) and CD235a ( HIR2 ) from BD and VISTA ( B7H5DS8 ) were purchased from eBioscience . Live dead staining kits were purchased from Invitrogen ( Carlsbad , CA , USA ) . Purified NA/LE Hamster anti-mouse CD3e ( 145-2C11 ) , CD28 ( 37 . 51 ) , phospho-Akt , and phospho-mTOR were purchased from BD Bioscience . Surface or intracellular staining were performed as we have described elsewhere [11 , 16] . TGF-β levels were measured following intracellular staining using Latency Associated Protein ( LAP ) . Paraformaldehyde fixed cells were acquired using a BD LSR Fortessa-SORP or Fortessa-X20 ( BD Bioscience ) . Data analysis was performed by using FlowJo software ( version 10 ) . Mice spleens were processed , and in order to obtain single-cell suspensions , spleen samples were ground between sterile frosted glass slides in 7 ml of 1× RBC lysis buffer and then filtered through nylon mesh . Cells were resuspended and processed in Dulbecco’s Modified Eagle’s Medium ( DMEM ) —high glucose ( Sigma ) with 1% penicillin/streptomycin ( Sigma ) , 10% FBS ( Sigma ) , and nonessential amino acids ( Sigma ) . For cell enrichment , CD71+ erythroid cells were enriched as we previously described elsewhere [10 , 11] with a purity typically exceeding 95% ( S2O Fig ) . Furthermore , isolated CD71+ erythroid cells were labeled with PE-conjugated anti-VISTA mAb ( MIH63 ) , followed by anti-PE microbeads ( Miltenyi Biotec ) , and passed through MACS separation columns ( Miltenyi Biotec ) . VISTA+ and VISTA− were isolated based on positive and negative selections , respectively . The purity of VISTA+ and VISTA− cells was approximately 95% as shown in S2P Fig . Naive CD4+ T cells were isolated from the adult mice spleens according to the manufacturing instruction ( Stemcell Technologies ) with purity > 90% ( S2Q Fig ) . Naïve CD4+ T cells were cultured in 96-well plates in the presence of either CD71+/VISTA+ or CD71+/VISTA− erythroid cells from WT mice and CD71+VISTA− from VISTA KO mice , supplemented with recombinant IL-2 ( 100 IU/ml ) , and stimulated with soluble anti-CD3 ( 3 μg/ml ) and anti-CD28 ( 1 μg/ml ) antibodies . Mouse recombinant TGF-β1 ( Biolegend ) and TGF-β1 blocker ( Sigma ) were also used as positive and negative controls , respectively . Cultures were analyzed 4–5 days later for FOXP3 induction . For functional assays , Tregs were isolated from either splenocytes of adult mice or following generation ( coculture of naïve CD4+ T cells with neonatal CD71+VISTA+ erythroid cells ) using Treg isolation kit ( Stemcell Technologies , Vancouver , Canada ) , then cocultured with effector T cells at different ratios for proliferation assay using CFSE dye ( Thermo Fisher Scientific , Waltham , MA , USA ) . Naïve CD4+ T cells were cultured for 24 h in 96-well plates with or without CD71+ erythroid cells in the presence of recombinant IL-2 ( 100 IU/ml ) and anti-CD3 ( 3 μg/ml ) and anti-CD28 ( 1 μg /ml ) . Then , phospho-AKT and phospho-mTOR were performed according to the manufacturing instruction . Enriched CD71+ erythroid cells were subjected to total RNA extraction in TRIZOL reagent ( Invitrogen ) using the RNeasy kit ( Qiagen , Venlo , The Netherlands ) . A Nano-Drop ND-1000 Spectrophotometer ( NanoDrop Technologies , Wilmington , DE , USA ) was used to check the quantity and quality of RNA for each sample . Usually , 500 ng of the isolated RNA was reverse transcribed employing a QuantiTect Reverse Transcription kit ( Qiagen ) . Gene expression of PD-1H ( VISTA ) , TGF-β , and TLRs was calculated by the 2−ΔΔCt method . Glyceraldehyde phosphatidyl hydrogenase ( GAPDH ) was used as a housekeeping gene for normalization of the cDNA levels . Samples collected at day 1 postnatal were used as the calibrator samples . The negative controls contained water or reverse-transcription negative RNA instead of template DNA . RNAseq libraries were constructed from 500 ng of total RNA using the TruSeq RNA Library Prep kit v2 ( Illumina , San Diego , CA , USA ) according to the manufacturer’s instructions at The Applied Genomic Core ( TAGC ) , University of Alberta . Libraries were sequenced on a NextSeq 500 instrument ( Illumina ) using a 75-bp paired-end protocol at an approximate depth of 12 M paired-end reads per sample . Transcripts abundance was quantified using Kallis [58] and 100 bootstraps . Differential expression analysis was conducted using Sleuth [59] . Data postprocessing was carried out with in-house R scripts . Statistical comparison between various groups was performed by the t test using PRISM software . In the gene expression assay , differences between adult group and newborns at different time points were evaluated using one-way ANOVA , followed by Tukey’s test for multiple comparisons . Results are expressed as mean ± SEM . p value < 0 . 05 was considered as statistically significant .
|
The primary role of the red blood cells is to transport oxygen , but we know relatively little about the other functions they perform . Following maturation , red blood cells exit the bone marrow and enter blood circulation . Their immature counterparts are normally absent or in very low frequency in the blood of healthy adults . However , we showed previously that immature red blood cells are abundant in the spleens of neonatal mice and in human umbilical cord blood and that these cells possess immunological properties . In this report , we studied a subset of neonatal immature red blood cells that express a protein called V-domain Immunoglobulin ( Ig ) Suppressor of T Cell Activation ( VISTA ) on their surface . We found that the presence of VISTA enables the cells to repeatedly produce the regulatory cytokine TGF-β . TGF-β induces a subset of naïve lymphocytes—the CD4+ T cells—and converts them into regulatory T cells , also known as Tregs . Tregs modulate and suppress other immune cells . Our studies provide novel insights , to our knowledge , into the immunological role of immature red blood cells in newborns .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immune",
"physiology",
"spleen",
"immunology",
"developmental",
"biology",
"antibodies",
"immune",
"system",
"proteins",
"white",
"blood",
"cells",
"animal",
"cells",
"proteins",
"gene",
"expression",
"t",
"cells",
"toll-like",
"receptors",
"biochemistry",
"signal",
"transduction",
"cell",
"biology",
"physiology",
"tgf-beta",
"signaling",
"cascade",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"regulatory",
"t",
"cells",
"immune",
"receptors",
"neonates",
"cell",
"signaling",
"signaling",
"cascades"
] |
2018
|
CD71+VISTA+ erythroid cells promote the development and function of regulatory T cells through TGF-β
|
Diagnosis of leptospirosis by the microscopic agglutination test ( MAT ) or by culture is confined to specialized laboratories . Although ELISA techniques are more common , they still require laboratory facilities . Rapid Diagnostic Tests ( RDTs ) can be used for easy point-of-care diagnosis . This study aims to evaluate the diagnostic performance of the RDTs LeptoTek Dri Dot , LeptoTek Lateral Flow , and Leptocheck-WB , prospectively . During 2001 to 2012 , one or two of the RDTs at the same time have been applied prior to routine diagnostics ( MAT , ELISA and culture ) on serum specimens from participants sent in for leptospirosis diagnosis . The case definition was based on MAT , ELISA and culture results . Participants not fulfilling the case definition were considered not to have leptospirosis . The diagnostic accuracy was determined based on the 1st submitted sample and paired samples , either in an overall analysis or stratified according to days post onset of illness . The overall sensitivity and specificity for the LeptoTek Dri Dot was 75% respectively 96% , for the LeptoTek Lateral Flow 78% respectively 95% , and for the Leptocheck-WB 78% respectively 98% . Based on the 1st submitted sample the sensitivity was low ( 51% for LeptoTek Dri Dot , 69% for LeptoTek Lateral Flow , and 55% for Leptocheck-WB ) , but substantially increased when the results of paired samples were combined , although accompanied by a lower specificity ( 82% respectively 91% for LeptoTek Dri Dot , 86% respectively 84% for LeptoTek Lateral Flow , and 80% respectively 93% for Leptocheck-WB ) . All three tests present antibody tests contributing to the diagnosis of leptospirosis , thus supporting clinical suspicion and contributing to awareness . Since the overall sensitivity of the tested RDTs did not exceed 80% , one should be cautious to rely only on an RDT result , and confirmation by reference tests is strongly recommended .
Leptospirosis is caused by microorganisms of the genus Leptospira . It is one of the world's most wide-spread zoonoses , with a mean global incidence of endemic and epidemic leptospirosis of 5 per 100 , 000 and 14 per 100 , 000 population , respectively [1] . It causes an acute febrile illness [2] with a wide diversity of milder clinical signs such as headache , malaise , myalgia , conjunctival suffusion and sometimes a transient rash . However , the illness can rapidly develop into a severe , potentially fatal form with a high mortality rate [3] . Leptospirosis is often overlooked since it mimics many other diseases , including dengue , malaria , influenza and hantavirus infections [4] , making differential diagnosis very difficult based on clinical grounds alone . Laboratory tests are therefore the basis of a confirmed case of leptospirosis . The most commonly used laboratory tests are based on detection of antibodies against the leptospires . Pathogenic leptospires enter the body through small cuts or abrasions , or via mucous membranes and possibly through wet skin . After infection , leptospires circulate in the blood stream , with a bacteremic phase lasting for up to 10 days post onset of the disease ( DPO ) . Detectable antibodies appear in the blood about 5–10 DPO [5] , and sometimes later , especially if antibiotic treatment is instituted [4] . These antibodies can be detected by a variety of laboratory assays such as the microscopic agglutination test ( MAT ) , enzyme-linked immunosorbent assay ( ELISA ) and indirect fluorescent antibody test ( IFAT ) [6] . Currently , the MAT is considered the reference standard in serodiagnosis and as such has a worldwide application . However , MAT and ELISA are technically demanding and relatively expensive tests and therefore not widely applicable in peripheral healthcare facilities , especially in tropical and subtropical developing regions where leptospirosis is most endemic . Culturing leptospires out of blood provides proof of infection but is insensitive [7] and has little clinical value for patient management as it can take weeks to months to confirm results . PCR on blood has proven to be useful in the first week of the disease [8] , however many laboratories are not equipped to run PCR tests . Hence , for most clinical situations rapid diagnostic tests ( RDTs ) can play an important role in immediate case detection and clinical management . Most commonly used RDTs are based on the immunochromatographic lateral flow technology . To date , a variety of RDTs have been described and evaluated in various papers [9]–[11] , most of these being short term retrospective evaluations and often concern evaluations of a single RDT . Our aim was to evaluate the diagnostic accuracy of three RDTs , applied on serum specimens from suspected leptospirosis patients from The Netherlands in a prospective cohort of leptospirosis suspected patients . Additional aims were to assess whether there are differences between the three tests and whether using the tests at different times since patient's onset of symptoms leads to differences in diagnostic accuracy . The RDTs used in this research were available in the Netherlands , or could be easily imported .
The Royal Tropical Institute ( KIT ) , Biomedical Research houses the WHO/FAO/OIE and National Collaborating Centre for Reference and Research on Leptospirosis ( NRL ) , which confirms about 99% of the suspected cases of leptospirosis in The Netherlands . The typical annual number of suspected cases is around 500 , of which approximately 30 are confirmed leptospirosis cases . About 50% of the confirmed cases have contracted leptospirosis during travel abroad . In the period of evaluation , July 2001 to August 2012 , the population in The Netherlands was stable at about 16 million . During this period , all human blood specimens sent by physicians practicing in the Netherlands to NRL for leptospirosis diagnosis were tested upon arrival by routine diagnostics . In most cases only one sample was received per participant , in other cases two or more samples . Further inclusion and exclusion criteria of samples and participants are depicted in a flow diagram ( Figure 1 ) . Laboratory tests routinely performed are MAT and in-house IgM-ELISA . Culture was done as described below . A single or a combination of two RDTs were prospectively performed for evaluation purposes . Patients were considered as having leptospirosis based on one or more of the following criteria: ( i ) single MAT titer with a pathogenic strain ≥1∶160 , ( ii ) single IgM-ELISA titer ≥1∶160 , ( iii ) positive culture or ( iv ) seroconversion/≥four-fold titer rise MAT or IgM ELISA ( titer ≤1∶20 to ≥1∶80 ) in paired samples taken at least 2 days apart [13] . The treating physician was encouraged to send multiple samples for laboratory testing for all participants . RDTs were applied prior to and independent of routine diagnostic testing . All tests were performed by skilled staff of NRL ( 10 persons ) who followed detailed protocols about interpretation of tests . NRL is accredited based on ISO 15189 since 2006 . All serological tests were performed on serum specimens which were inactivated in a 56°C water bath for 30 minutes before testing . Data were entered into a Laboratory Information System ( LASSIST , Mechatronics Software Applications BV , the Netherlands ) and exported and analyzed in SPSS ( version 19 , IBM , NY , USA ) . These included patient data obtained from the request form ( i . e . gender , date of birth , date of onset , travel history ) . The results of each diagnostic test of every sample were entered into the database . Follow-up samples taken less than two days after the first sample were excluded . Indeterminate results were regarded as negative , unless otherwise stated . This data collection was exempted from ethical review of human subjects research by the Medical Ethical Review Committee of the Academic Medical Centre , University of Amsterdam ( W12_076#12 . 17 . 0092 ) . All data presented have been de-identified and were not attributable to individual patients .
The overall sensitivity and specificity , calculated on all samples from early acute till the late convalescent phase showed a sensitivity of 75% ( 95% CI 69% to 79% ) for LeptoTek Dri Dot , 78% ( 95% CI 69% to 85% ) for LeptoTek Lateral Flow and 78% ( 95% CI 71% to 83% ) for Leptocheck-WB . The specificity was 96% ( 95% CI 95% to 97% ) for LeptoTek Dri Dot , 95% ( 95% CI 94% to 96% ) for LeptoTek Lateral Flow and 98% ( 95% CI 97% to 98% ) for Leptocheck-WB ( Table 3 ) . There were no marked differences between the three tests; the sensitivities and specificities were similar with overlapping confidence intervals . When considering only the first sample that was sent in for each patient , the sensitivity of each test dropped dramatically from 75% to 51% and from 78% to 55% for the LeptoTek Dri Dot and the Leptocheck-WB , respectively . The sensitivity of the LeptoTek Lateral Flow decreased from 78% to 69% , although not a statistically significant change . The specificity of all tests remained more or less the same . Test results from paired samples ( either one of the samples positive ) increased the sensitivity significantly from 51% to 82% for the LeptoTek Dri Dot and from 55% to 80% for the Leptocheck-WB . The increase from 69% to 86% for the LeptoTek Lateral Flow was not statistically significant . The corresponding decrease in specificity was significant , i . e . from 96% to 91% for the LeptoTek Dri Dot , from 96% to 84% for the LeptoTek Lateral Flow and from 98 to 93% for the Leptocheck-WB ( Table 3 ) . For 2733 participants ( 53 . 1% of study participants ) the first day of onset of symptoms was known . All three tests show a lower sensitivity during the early acute phase of the disease ( till DPO 4 ) , which increased during DPO 5–10 and DPO 11–20 , while the specificity of all tests remained relatively stable ( Table 4 ) . LeptoTek Lateral Flow was performing the best at DPO 0–4 ( sensitivity of 62% , 95% CI 41% to 79% and specificity of 98% , 95 CI 93% to 99% ) . The proportion of the indeterminate results for the 1st sample for LeptoTek Dri Dot were 10/256 ( 4% ) in the participants fulfilling the case definition and 85/2903 ( 3% ) in the participants not fulfilling the case definition . For the LeptoTek Lateral Flow , these proportions were 4/108 ( 4% ) , respectively 173/1292 ( 13% ) , and for the Leptocheck- WB 17/183 ( 9% ) , respectively 239/2551 ( 9% ) . Allocation of indeterminate results to positive scores did not substantially change sensitivity , but it did have an impact on specificity ( Figure 2 ) . For the LeptoTek Dri Dot , the specificity decreased from 96% to 93% for the 1st submitted sample and from 91% to 81% for the paired samples . The LeptoTek Lateral Flow showed a decrease of the specificity from 96% to 82% for the 1st sample and 84% to 62% for the paired samples , while the Leptocheck-WB showed a decrease from 98% to 88% and from 93% to 80% respectively . About 28% of the participants with an initial indeterminate result for the LeptoTek Dri Dot and Leptocheck-WB were later confirmed with leptospirosis in follow-up testing , and had a positive RDT compared to only 10% of participants with an initial negative result . For the LeptoTek Lateral Flow , the numbers are somewhat different with 9% positive results after the first sample was indeterminate , and 4% positive results after the first sample was negative , but the same trend is present ( Table S4 ) . Exclusion of indeterminate results showed an increasing sensitivity and decreasing specificity for all RDTs and for all time points , though not statistically significant . When stratifying the samples according to the defined time-periods of the disease , the same trend was observed ( Table S3 ) . The sensitivity of RDTs appeared to depend on the infecting serogroup ( Table S3 ) . In general infecting serogroup Icterohaemorrhagiae yielded a higher sensitivity for all three RDTs compared to the other categories of serogroups . Differences were significant in the following cases: The LeptoTek Dri Dot showed a higher sensitivity for the paired samples in the Icterohaemorrhagiae infections ( 98% ) compared to the other infections ( 81% ) and non-classifiable serogroup infections ( 60% ) . The 1st submitted samples of the latter category also had a lower sensitivity ( 38% ) compared to the Icterohaemorrhagiae group infections ( 67% ) . The LeptoTek Lateral Flow showed a higher sensitivity in both the 1st submitted samples and the paired samples for the Icterohaemorrhagiae infections ( 85% respectively 100% ) compared to for ‘non-classifiable serogroups’ ( 51% respectively 63% ) . Leptocheck-WB showed a higher sensitivity in the 1st samples ( 68% ) as well as the paired samples ( 95% ) for the Icterohaemorrhagiae infections compared to the category ‘non-classifiable serogroups’ ( 1st submitted sample 38% , paired samples 65% ) . To investigate the consistency of the diagnostic accuracy of these RDTs over the time period 2001 to 2011 , the diagnostic accuracy based on the 1st submitted sample and paired samples for each year for each test was compared ( Figure 3 ) . Significant variation was observed in the following cases: for the 1st sample submitted , the sensitivity of the LeptoTek Dri Dot decreased from 77% in 2001 to 37% in 2005 combined with increasing specificity from 93% to 98% . During the same years the paired samples showed a decrease in sensitivity from 100% to 67% . Also the LeptoTek Lateral Flow showed on the 1st submitted sample a decreasing sensitivity from 100% in 2001 to 50% in 2003 , whereas the specificity increased from 87% to 99% . For the paired samples , the specificity increased from 60% to 100% . On the contrary , based on the 1st submitted sample the Leptocheck-WB showed an increase in sensitivity , from 36% in 2005 to 78% in 2009 , combined with a decreasing specificity from 100% to 97% .
The LeptoTeK Lateral Flow presents in all scenarios with the best sensitivity and equally good specificity of all three RDT tests . All three tests , LeptoTek Dri Dot , LeptoTek Lateral Flow and Leptocheck-WB present useful antibody tests contributing to the diagnosis of leptospirosis . For sure , confirmation of clinical suspicion will contribute to increased local awareness of leptospirosis . Confirmation might also be beneficial for the clinical management of the patient . On the other hand , it should be noted that , especially in the early phase , a negative RDT and a high clinical suspicion still warrants antibiotic treatment since ( untreated ) leptospirosis is a potential fatal disease . Unfortunately , currently LeptoTek Dri Dot and LeptoTek Lateral Flow are not available due to manufacturer issues , presently leaving few options . The overall sensitivity of the tested RDTs did not exceed 80% , while their performance might depend on batch-to-batch and year-to-year variations as well as on varying ecological niches containing different circulating serovars . This latter drawback might be extended with a reduced diagnostic accuracy due to past leptospiral infections or infections with other causative agents in high endemic areas , causing cross-reactions in these tests [9] . For these reasons , one should be cautious to only rely on an RDT result . Confirmation by reference tests is strongly recommended , and further conclusive studies are needed in endemic regions . From this study we have seen that rapid testing is not synonymous with easy testing . Reading of tests by eye is subjective and depends on the experience of the reader . At least it is of great importance that a test result , in case of doubt , is reported as such , indicating the need for a follow-up sample , especially evading the inclination of the reader to score a doubtful signal as a positive score .
|
Leptospirosis is one of the world's most spread zoonoses causing acute fever . The illness can rapidly develop into a severe , potentially fatal , form with a high mortality rate . Laboratory tests are needed to confirm the diagnosis . Culturing leptospires from patient material can take months to grow . Therefore , most used laboratory tests are based on detection of antibodies against leptospires . The microscopic agglutination test is considered the reference standard but is only performed at specialized laboratories . In this study , we measured the diagnostic accuracy of three rapid diagnostic tests ( RDTs ) by doing a prospective evaluation during 11 years . These tests produce results within 15 minutes . The overall sensitivities ( 77% ) and specificities ( 96% ) were similar for the RDTs . Evaluating the first submitted specimen resulted in lower sensitivities ( 51% for LeptoTek Dri Dot , 69% for LeptoTek Lateral Flow , and 55% for Leptocheck-WB ) . When paired specimens were evaluated , the sensitivity increased although the specificity decreased ( 82% respectively 91% for LeptoTek Dri Dot , 86% respectively 84% for LeptoTek Lateral Flow , and 80% respectively 93% for Leptocheck-WB ) . Based on these results confirmation by reference tests is still strongly recommended , although the RDTs contribute to the diagnosis of leptospirosis , thus supporting clinical suspicion and contributing to awareness .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"bacterial",
"diseases",
"infectious",
"diseases",
"immunoassays",
"neglected",
"tropical",
"diseases",
"immunology",
"biology",
"immunologic",
"techniques",
"leptospirosis"
] |
2013
|
Prospective Evaluation of Three Rapid Diagnostic Tests for Diagnosis of Human Leptospirosis
|
A Schistosoma mansoni cercarial antigen preparation ( cercarial transformation fluid – SmCTF ) was evaluated for detection of anti-schistosome antibodies in human sera in 4 collaborating laboratories . The performance of SmCTF was compared with that of S . mansoni egg antigens ( SmSEA ) in an indirect enzyme-immunoassay ( ELISA ) antigen assay , the latter being used routinely in 3 of the 4 participating laboratories to diagnose S . mansoni and S . haematobium infections . In the fourth laboratory the performance of SmCTF was compared with that of S . japonicum egg antigens ( SjSEA ) in ELISA for detection of anti-S . japonicum antibodies . In all 4 laboratories the results given by SmCTF in ELISA were very similar to those given by the antigen preparation routinely used in the respective laboratory to detect anti-schistosome antibodies in human infection sera . In so far as the ELISA results from SmCTF are thus so little different from those given by schistosome egg antigens and also cheaper to produce , the former is a potentially useful new diagnostic aid for schistosomiasis .
More than 200 million people in over 70 countries world-wide are infected with schistosomes with infection-induced morbidity being particularly pronounced in sub-Saharan Africa [1] , [2] . Humans become infected as a result of swimming , bathing and fishing in water in which infected intermediate host snails have released free-swimming cercariae that can penetrate human skin . The heaviest schistosome infections are generally found in children and young adults and in recognition of this school children are the main target of schistosomiasis control programmes based on treatment with praziquantel . Prior to instigating control the prevalence and intensity of infection is generally estimated by microscopic detection of eggs in faecal or urine samples , which is a relatively slow and laborious process . Insensitivity is another serious defect of egg detection methods of diagnosis , especially of the intestinal schistosome infections [3] , [4] and many light infections are missed because of the absence of eggs in the small volumes of excreta that can be routinely examined microscopically [5]–[9] These limitations impose significant constraints on current control initiatives [10] , [11] Considerable effort has been expended in the effort to develop immunodiagnostic tests that are an improvement on microscopical parasitology . It has been argued that methods to detect circulating or excreted schistosome antigens are desirable because they are likely to reflect active infection most accurately . However , the sensitivity of antigen detection tests seems to be no better than that of microscopy , particularly with regard to detection of faecally-excreted eggs of S . mansoni and in situations in which low egg counts pertain [12] , [13] Antibody detection tests have often been deemed unsuitable for diagnosis of schistosomiasis , mainly because of their apparent lack of specificity and inability to distinguish active from inactive infection – namely the common observation that many subjects that are antibody-positive are egg-negative by microscopy . However , possible alternative explanations for the lack of specificity are that the many instances of antibody-positivity , egg-negativity reflect the failure of insensitive microscopy to detect eggs in subjects who are lightly-infected [3] or who have been treated with sub-curative drug doses [14] . Indeed it has been demonstrated that in some patients antibody levels do decline following treatment [15] , particularly antibodies against the soluble egg antigen fraction CEF6 and patients with more steeply declining anti-CEF6 antibody titres were considered to have been better cured than those with titres that remained higher [16] . There is of course also the possibility that antibody false-positives are due to heterologous infectious agents . Despite their failings , antibody-detection is for some time likely to remain the best available method for diagnosis in areas of low intensity of schistosome infection [11] , [17] . Tourists and other visitors to schistosome endemic areas who become infected with schistosomes commonly only have light infections and because praziquantel is such a safe drug travellers' medicine clinics now often base their treatment decisions on the result of an antibody-detection diagnostic test alone . Soluble S . mansoni egg antigens ( SmSEA ) in enzyme immunosorbant assay ( ELISA ) formats are frequently the diagnostic method of choice in these clinics [18]–[20] and such immunoassays can also give meaningful results in endemic areas [21] . SmSEA is however a relatively expensive commodity , requiring for its production the infection of relatively large numbers of laboratory animals ( generally mice or hamsters ) , from the tissues of which the eggs are isolated after the infections become patent . Diagnostic tests based on SmSEA would therefore not be ideal for deployment in tropical countries with endemic schistosomiasis , most of which are resource-poor . Based on the evidence that much of the antibody induced by S . mansoni infections is specific for glycanic epitopes [22] , [23] and that S . mansoni egg and cercarial molecules have many glycanic epitopes in common [24]–[26] , we have begun to investigate whether soluble S . mansoni cercarial antigens ( in the form of a preparation we have named ‘cercarial transformation fluid’ – SmCTF ) can be substituted for SmSEA as an antibody target in ELISA . A preliminary study has indicated that SmCTF did indeed give very similar reactivity to SmSEA when reacted in ELISA with sera from schistosome-infected patients who had attended a travellers' medicine clinic in London [27] . In that study the sera , when tested previously , all had reacted positively with SmSEA , but they were designated just as schistosome-infected i . e . , there was no differentiation between S . mansoni and S . haematobium infection . There was no difference between the rate of cross-reactivity of SmCTF with sera from patients with other infections than that of schistosomiasis compared to that achieved with SmSEA [27] . A second study has also demonstrated that SmCTF performs equivalently to SmSEA for diagnosis of S . mansoni infections in an endemic setting , and indicates that SmCTF may actually be more specific than SmSEA for diagnosis in endemic areas [28] . In this study we have extended that work in a collaborative study involving 3 different parasite diagnosis laboratories in Europe: respectively , in Scotland , England and The Netherlands , all of which deal with travellers and immigrants from Africa or the Middle East who may be infected with S . mansoni and/or S . haematobium . The species of schistosome infecting many of the serum donors could be determined by prior clinical investigations routinely carried out in the respective laboratories , including travel history , water exposure questionnaire and/or microscopy of urine and stools for schistosome ova . The opportunity has also been taken to test the reactivity of human sera from an area endemic for S . japonicum in The People's Republic of China . In all 4 settings the performance of SmCTF was compared directly with SmSEA in ELISA . In each of the four laboratories the ‘in-house’ procedures in routine use for performing ELISA were used to compare the antigen preparations , with therefore no effort having been made to co-ordinate or validate the methodology at this stage .
All the serum samples used in this study were routine diagnostic samples which had previously been sent to the respective laboratory from other hospitals/medical practitioners with the request they be tested for anti-schistosome antibodies . The samples were therefore not collected specifically for this study and were taken from stored collections of sera . In the case of the participating laboratory having collected the sample for a previous routine diagnosis , sample donors were asked if they had any objections to usage of the samples for future research . If no objection was provided , the samples were deemed suitable for use in this study: i . e . , a determination of the merits of different diagnostic tests . After processing every sample sent to one of the laboratories for testing was stored frozen in accordance with Standard Operating Procedures for “sample collection and storage” that are in place in each of the laboratories . All samples are maintained for a set period of time in case a sample requires retesting . The collections of sera so accrued provide reference material for research/development studies such as this . The S . japonicum infection sera used by Nanjing Medical University ( see below ) had been collected as part of the National Key Technologies R&D Program for China Tenth Five-Year Plan ( grant no . 2004BA718B05 ) . The project was reviewed and approved by the institutional review board of Nanjing Medical University . The other 3 serum collections used in this study had been given in-house names which were and are used in references to them; e . g . SPDRL Sera or Amsterdam Medical Centre Serum Bank . These were in turn parts of larger reference collections: Scottish Parasite Clinical Reference Collection , the AMC Parasitology Serum Bank and the LSTM Diagnostic Parasitology Laboratory Serum Bank . The regulations to which the participating laboratories adhered with regard to use of samples in their possession covered the use of these samples in this particular study . Application for IRB approval was deemed unnecessary and irrelevant because any sample which was received by any of the participating laboratories from elsewhere was given a unique identifier . No personal identifying data , such as name , date of birth , gender , and/or age were retained on the collection tube . All samples were stored in numerical order using a unique identification number to ensure they were anonymised for any subsequent study . Each sample used in this study was thus completely anonymised and the results were a ) not entered onto any computerised reporting system; b ) not accessible by any outside body , and c ) not reported to the donor or the sample sender and therefore not reported to the patient . In the participating laboratories it is standard practice for stored fully-anonymized serum samples to be used for evaluation of new diagnostic tests , as in this study , without the need for permission to be sought from regulatory bodies . Consent was not required for this particular study because it was simply a comparison of diagnostic methodologies using fully anonymised samples which were already part of collections of reference material . Each sample was being tested for the same parasitic infection for which it had initially been screened . No individual's results from this study have been reported elsewhere and no follow-up samples were necessary . The work on laboratory animals was approved by the Ethical Review Procedure of the University of Nottingham . Further details are available at this link: http://www . nottingham . ac . uk/animalresearch/erp/erp . aspx . The work was conducted according to the UK Animal ( Scientific Procedures ) act 1986 with personal and project licence authorities held by MD ( numbers PIL70/3255 and PPL40/3024 respectively ) . The study was conducted in accordance to Nottingham University's guidelines for animal husbandry which meet the UK Home Office Code of Practice for the Housing and Care of Animals used in Scientific Procedures . Further details are available at this link: http://www . nottingham . ac . uk/animalresearch/animalwelfare/animalwelfare . aspx . Four medical parasitology laboratories collaborated in this study to compare the performance of two S . mansoni antigen preparations in ELISA with sera from schistosome-infected people: The Scottish Parasite Diagnostic Reference Laboratory , Glasgow , Scotland ( SPDRL ) , The Diagnostic Parasitology Laboratory at the Liverpool School of Tropical Medicine ( LSTM ) , the Parasitology Section , Academic Medical Center , Netherlands ( AMC ) and the Department of Pathogen Biology , Nanjing Medical University , Nanjing , China ( NJMU ) . Two S . mansoni antigen preparations , SmCTF and SmSEA , were produced by BioGlab Ltd ( Nottingham , UK ) and distributed to the above-named 4 laboratories for comparison of their performance in ELISA . SmCTF was prepared from cercariae , the shedding of which was induced from Biomphalaria glabrata snails . The cercariae were shed into deionized water at 30°C under a 60 W tungsten light and the larval suspensions concentrated by suction over a glass fibre filter , followed by gravity-sedimentation in the dark in approximately 20 ml water at 4°C . Excess water was removed from over the sedimented cercarial pellets , the latter being generally of 1 ml volume or more . The pellets were resuspended in 5 ml phosphate buffered saline ( PBS , pH 7 . 2 ) and aspirated 15 times through a 20 gauge disposable syringe needle in 5 ml phosphate-buffered saline ( pH 7 . 2 ) using a 5 ml disposable plastic syringe [29] . Suspensions of dispersed , ‘transformed’ schistosomula and their now separate tails were incubated in plastic Petri dishes for 45 min at room temperature ( RT ) . After incubation schistosomula and tails were removed by centrifugation ( 5 min , 250× g ) and the supernatant ( SmCTF ) was collected and stored at −80°C . The protein concentration of SmCTF was approximately 1 . 0 mg per ml; batches were aliquoted in 1 ml volumes , freeze-dried and distributed to the 4 collaborating laboratories for use in this study . Extracts of S . mansoni soluble egg antigens ( SmSEA ) were prepared as described earlier [30] . S . japonicum egg antigens ( SjSEA ) were prepared with eggs taken from rabbits 45 days after they had been infected with approximately 1 , 000 S . japonicum cercariae . Livers were removed and kept at 4°C overnight in order to better isolate eggs . Eggs were harvested from the homogenized livers by differential centrifugation . The absence of contaminating rabbit tissue fragments in the egg preparation was checked by microscopic analysis . Purified eggs were suspended in PBS and homogenized on ice . After repeated freezing and thawing , the homogenate was centrifuged at 13 , 800 g at 4°C for 20 min , and the supernatant was used as S . japonicum soluble egg antigen ( SjSEA ) . All data were analysed using GraphPad Prism 4 and GraphPad InStat 3 ( GraphPad Software , USA ) . Correlations were calculated using either Spearman's r correlation or Pearson's r correlation dependent upon the normality of the data . Sensitivities and specificities were calculated using the SmSEA-ELISA as the ‘gold-standard’ test unless elsewhere stated . Statistical differences in mean SmSEA OD readings and SmCTF OD readings were calculated using a t-test or the Wilcoxon matched pairs test as appropriate . Statistical differences between mean OD readings of one species compared to another were calculated using either Welch corrected unpaired t-tests or Mann Whitney U tests . Significance was assigned at p<0 . 05 .
Figure 1 shows the correlation between respective anti-SmCTF and anti-SmSEA OD450 readings from sera of 59 school pupils who attended a training course in the vicinity of the White Nile in Uganda . There was 91% correlation ( 95% CI 85 . 1–94 . 6; p<0 . 0001 ) . It is apparent that the majority of individual anti-SmCTF OD450 readings ( mean = 0 . 71 , SD = 0 . 53 ) are greater than the respective OD450 anti-SmSEA OD450 values ( mean = 0 . 50 , SD = 0 . 47 ) ; a difference which is significant ( p<0 . 0001 ) . The sensitivity of the SmCTF compared with the SmSEA was 97 . 1% and the specificity was 80% . The positive predictive value was 86 . 8% and the negative predictive value was 95 . 3% . Figure 2 shows the anti-SmCTF and anti-SmSEA OD450 readings for those of the 59 subjects in Figure 1 who had anti-SmSEA ODs of between 0 and 0 . 3 . The cut-off values for SmCTF and SmSEA , estimated from the ELISA results of 50 negative controls , were respectively 0 . 25 for SmCTF and 0 . 22 for SmSEA and have been indicated in the graph . Five sera gave ODs which were above the SmCTF cut-off , but below the SmSEA cut-off; i . e . , recorded a positive result with the former antigen , negative with the latter , a result consistent with SmCTF being more reactive than SmSEA with S . mansoni infection sera in this laboratory . Figure 3 shows individual SmCTF OD450 values plotted against the respective SmSEA OD450 values of sera from 8 subjects known to be infected with S . haematobium by virtue of the presence of eggs in urine and also from 8 school children who had visited Lake Malawi , and thereby assumed also to have been exposed to S . haematobium . Two of the 8 children were positive for both antigens and 6 were negative , the latter showing as a closely knit group in the lower left corner of the figure . There was 90 . 8% correlation between the SmCTF OD450 values and the SmSEA OD450 values ( 95% CI 74 . 3–96 . 9; p<0 . 0001 ) , and no significant difference between their means . Figure 4 plots the individual results of the respective anti-SmCTF above cut-off value OD450 results from the S . mansoni-infection sera in Figure 1 and S . haematobium-infection sera in Figure 3 . The mean of the latter values was marginally higher , but not significantly so . In contrast to the results in Figure 4 , LSTM found that 12 S . mansoni infection sera gave a higher mean OD405 reading with SmCTF than 40 S . haematobium infection sera ( Figure 5 ) , a difference that was significant ( p<0 . 02 ) . When the reactions of individual sera against SmCTF were plotted against their respective reactions against SmSEA ( Figure 6 ) the anti-SmCTF OD405 values of the majority of the S . mansoni infection sera appeared to be higher than their respective anti SmSEA OD405 values ( although this difference is not significant ) , while the reactivities of the S . haematobium infection sera were relatively evenly distributed about the equivalence line ( Figure 7 ) . Similarly , sera from a group of subjects suspected to be infected with schistosomes , but not determinable as either S . mansoni or S . haematobium , reacted relatively similarly to both SmCTF and SmSEA in ELISA ( Figure 8 ) . For the S . mansoni infection sera , there was 88 . 9% correlation between the anti-SmCTF OD405 values and the anti-SmSEA OD405 values ( 95% CI 64 . 4–96 . 9; p = 0 . 0001 ) . For the S . haematobium infection sera , there was 77 . 4% correlation between the anti-SmCTF OD405 values and the anti-SmSEA OD405 values ( 95% CI 60 . 4–87 . 7; p<0 . 0001 ) . For the suspected cases , there was 85 . 4% correlation ( 95% CI 73 . 4–92 . 2; p<0 . 0001 ) . Using either egg microscopy or the SEA-ELISA as the ‘gold-standard’ diagnostic test , the CTF-ELISA had a sensitivity of 100% to S . mansoni infections as there were no false-negative results . The SEA-ELISA gave one false-negative result with S . mansoni infection sera and therefore had a sensitivity of 91 . 6% . Again using egg microscopy as the ‘gold-standard’ , the CTF-ELISA had a sensitivity of 87 . 5% to S . haematobium infections with 5 false-negative results . Two of these were very close to the cut-off value of 0 . 200 ( 0 . 198 and 0 . 199 ) , and if these are taken to be positive results sensitivity is increased to 92 . 5% . One of the remaining false-negatives was with serum from a patient with mainly calcified eggs in the urine; this serum sample also tested negative in the SEA-ELISA . The SEA-ELISA had a sensitivity of 92 . 5% with 3 false-negative results . One of these false negatives gave an OD reading that was close to the cut-off value ( 0 . 188 ) . Using the SEA-ELISA as the gold-standard test , the CTF assay had a sensitivity of 91 . 9% or 97 . 3% if the two-values that were close to the cut-off are taken to be positive results . All of the sera from cases of suspected schistosomiasis gave a positive result with at least one of the antigens in ELISA . Using SEA in ELISA as the gold standard , the CTF-ELISA had a sensitivity of 97% with only one false-negative result . There were two false-positive results , which could reflect the CTF being more sensitive than SEA . Figure 9 shows SmCTF OD492 values plotted against the respective SmSEA OD492 values of a total of 46 sera; 21 from patients with parasitologically-proven Schistosoma infection , 20 sera that were positive in routine serology and 5 negative controls . These differ somewhat from the results of the SPDRL and LSTM; in particular a few had high SmSEA ODs and low SmCTF ODs ( see ringed points in figure 9 ) . Three of these points are the results from patients with parasitologically-proven S . mansoni infection ( two of which gave negative SmCTF-ELISA results ) ; the other is not associated with any one infection species or status . Despite these outlying results , there was 72 . 45% correlation ( 95% CI 54 . 4–84 . 1; p<0 . 0001 ) . In contrast to the SPDRL's results , the majority of individual anti-SmSEA OD readings ( mean = 0 . 534 , SD = 0 . 114 ) are greater than the respective anti-SmCTF OD values ( mean = 0 . 365 , SD = 0 . 152 ) for S . mansoni infection sera , a difference which is significant ( p = 0 . 0124 ) . Also , in contrast to the results from the SPDRL and LSTM ( figures 4 and 5 ) , the AMC found that 5 S . haematobium-infection sera gave a higher mean OD reading with SmCTF ( mean = 0 . 49 , SD = 0 . 15 ) than 10 S . mansoni infection sera ( mean = 0 . 36 , SD = 0 . 04 ) , a difference that was significant ( p<0 . 05 ) . Sera from 42 subjects living in the vicinity of Poyang Lake , Jiangxi Province , China , an area endemic for S . japonicum , and from 20 volunteers with no history of living in a schistosome endemic area were tested for reactivity in ELISA against 3 antigens: SmCTF , SmSEA and SjSEA , the last of these three being the antigen routinely used to detect anti-S . japonicum antibodies in this laboratory . The individual OD450 results for the 3 groups of sera against the 3 different antigens are shown in Figure 10 . The mean reactivities of each of the 3 groups against each antigen are very similar . ( Egg-positives: SmSEA mean = 1 . 28 , SD = 0 . 20; SjSEA mean = 1 . 29 , SD = 0 . 20; SmCTF mean = 1 . 24 , SD = 0 . 22; egg negatives: SmSEA mean = 0 . 91 , SD = 0 . 24; SjSEA mean = 0 . 91 , SD = 0 . 24; SmCTF mean = 0 . 87 , SD = 0 . 26; negative controls: SmSEA mean = 0 . 37 , SD = 0 . 08; SjSEA mean = 0 . 35 , SD = 0 . 08; SmCTF mean = 0 . 37 , SD = 0 . 07 ) . The 3 respective groups of sera gave very similar OD450 values irrespective of whether an egg antigen preparation from either S . japonicum ( mean = 0 . 88 , SD = 0 . 44 ) or S . mansoni ( mean = 0 . 88 , SD = 0 . 43 ) , or the S . mansoni cercarial antigen preparation ( mean = 0 . 85 , SD = 0 . 42 ) was the target in ELISA . The correlation between the anti-SjSEA OD450 values and the anti-SmCTF OD450 values was 96 . 6% ( 95% CI 94 . 2–98 . 0; p<0 . 0001 ) . There was 96 . 4% correlation between the anti-SmSEA OD450 values and the anti-SmCTF OD450 values ( 95% CI 94 . 0–97 . 9 ( p<0 . 0001 ) , and 98 . 5% correlation between the anti-SjSEA OD450 values and the anti-SmSEA OD450 values ( 95% CI 97 . 5–99 . 1; p<0 . 0001 ) . The OD450 values that discriminated between a +ve and −ve outcome for each serum in this experiment were calculated as the mean+2× SD of the uninfected control OD450 values of the uninfected sera . The cut-off values thus calculated were for: SmSEA = 0 . 539 , SmCTF = 0 . 498 and SjSEA = 0 . 502 . All three of the antigens used in ELISA gave positive results on the sera from egg-positive cases . Most of the sera from endemic egg-negative cases tested positive with all three antigens in ELISA , i . e . gave false-positive results compared to egg microscopy . All of the negative controls gave negative ELISA results with the three antigens , except for one which gave a false-positive result with SjSEA . Using SjSEA in ELISA as the ‘gold standard’ for diagnosis of S . japonicum infection , SmCTF had 95 . 24% sensitivity and 100% specificity . The two ‘false-negatives’ seen were due to positive results with SjSEA on a negative control and an egg-negative case . SmSEA gave exactly the same positive and negative results as SmCTF , and so also had 95 . 24% sensitivity and 100% specificity compared to SjSEA . SmSEA and SmCTF had a positive predictive value ( PPV ) of 100% and a negative predictive value ( NPV ) of 90 . 9% . Using egg microscopy as the gold standard , SmCTF had a sensitivity of 100% and a specificity of 62 . 5% . 15/17 of the endemic area egg-negative cases gave positive results with SmCTF in ELISA . There were no false-negatives . SmSEA had the same sensitivity and specificity values as SmCTF . SjSEA in ELISA compared with egg microscopy had a sensitivity of 100% and was 59 . 5% specific; i . e . , less specific than SmCTF and SmSEA as there were 2 more false-positives using SjSEA . There were no false-negatives . SmSEA and SmCTF had a PPV of 62 . 5% and a NPV of 100%; SjSEA had a PPV of 59 . 5% and a NPV of 100% . A flow-chart summarizing the results from the 5 laboratories is given in supporting information Figure S1 .
The above results indicate that in three different laboratory settings the novel SmCTF antigen preparation is as effective in detecting anti-S . mansoni , anti-S . haematobium and anti-S . japonicum antibodies as SmSEA . The latter antigenic preparation , derived from S . mansoni eggs , has been routinely used to diagnose schistosome infections in the 3 European laboratories participating in this study , all of which have diagnostic service responsibilities . There were some minor differences between the 3 European laboratories , perhaps a reflection of the way that each performs its ELISA reactions and incidentally an indication that there is as yet no standardized and widely used method for detecting anti-schistosome antibodies . The results from the SPDRL show that for the suspected S . mansoni-infection sera the sensitivity of SmCTF relative to SmSEA is high with only one false-negative in 59 results . Specificity is not so high: 80% as a result of 5 false-positives , but this could be due to failure of the SmSEA to detect light infections [3] . It could however alternatively be due to sensitisation by cercarial antigens of bird and animal schistosome species , though an association between these and the White Nile in Uganda has seldom if ever been reported . Indeed , cross-reactivity between SmCTF and antibodies from cercarial dermatitis patients will need to be investigated before this assay is fully validated . For the S . haematobium-infection sera examined in SPDRL SmCTF was 100% sensitive and 100% specific as there were no false-negatives or false-positives relative to SmSEA . The AMC results differed from those of the other two European laboratories since those infection sera with higher SmSEA OD readings had lower SmCTF OD readings and three of them were parasitologically-proven S . mansoni-infection sera . Two of these gave false-negative results . The sample size here was , however , considerably smaller than that of the other two laboratories with only 10 S . mansoni-infection sera tested . At the SPDRL , SmCTF gave higher OD readings with S . mansoni-infection sera than SmSEA ( p<0 . 05 ) , but this was not the case in the other two laboratories; the LSTM's results showed little distinction between SmCTF and SmSEA with regard to the OD values given by individual sera in ELISA , and at the AMC SmSEA gave higher OD readings than SmCTF with S . mansoni infection sera ( p<0 . 02 ) . Again there is the problem of a small sample size with the results from the AMC . It is perhaps surprising that the S . haematobium-infection sera reacted well with the S . mansoni antigen preparations in all three laboratories , both cercarial-derived ( SmCTF ) and egg-derived ( SmSEA ) . At the AMC they reacted better with SmCTF than the S . mansoni-infection sera ( p<0 . 05 ) , but at the LSTM S . mansoni sera reacted better than S . haematobium sera ( p<0 . 02 ) and there was no significant difference between the two at the SPDRL . It has long been known that anti-cercarial antibodies in human infection sera can be demonstrated , for example by means of the so-called cercarienhüllen reaction [33] , but schistosome cercariae have not often been used to detect antibodies in immunodiagnosis of schistosomiasis , perhaps because of a relative lack of sensitivity and specificity [3] , [34] . Higher anti-larval:anti-adult antibody levels may be useful in discriminating between acute and chronic infection [35] . Two more recent studies have indicated that the novel cercarial antigen SmCTF performs equivalently to SmSEA for the diagnosis of both S . mansoni and S . haematobium infections [27] , [28] . It is suspected that the anti-schistosome antibodies that were detected by the SmCTF-ELISA had been induced by immunogens derived from eggs lodged in the tissues of infected humans ( rather than by the schistosome cercariae or worms that caused the infection ) . Furthermore , the carbohydrate molecules derived from the cercarial glycocalyx and which become solubilised during the transformation of cercariae to schistosomula are suspected to be the principal antigenic moiety in SmCTF reactive here in ELISA . The serological cross-reactivity indicated by the present results , both between schistosome species and between antigens derived from schistosome cercarial and egg stages , is likely to be due to the presence of the same carbohydrate epitopes on the molecular constituents of eggs and larvae of the different species [22] , [23] , [36]–[38] . In view of the likely high dependence of the anti-cercarial antigen antibody reactivity being induced by egg antigens , it seems improbable that SmCTF will be of use in detecting pre-patent or single-sex worm-alone infections . It seems to us less likely that protein epitopes are involved to any significant extent in the antigen/antibody reactions described here because the reactivity of human infection sera against SmCTF in Western immunoblots is negated by prior treatment of the antigens with sodium metaperiodate solution ( Francklow and Doenhoff , unpublished result ) . Furthermore , antibodies reactive against one of the principal protein constituents of SmCTF , the cercarial elastase , seem not to be present in human infection sera [39] . Many of the schistosome-infected subjects encountered by the three European laboratories involved in this study during the course of their diagnostic service responsibilities have light levels of infection which routine parasitological microscopy and antigen-detection methods often are too insensitive to detect . Despite their disadvantages , antibody-detection methods have therefore to be resorted to in the absence of any other useful test to detect infection directly . Because it is such a safe drug , praziquantel is then often automatically administered to those who are antibody-positive , despite a risk of some of the antibody-detection diagnostic test results being falsely-positive with regard to active infection . Praziquantel has become the basis for control of endemic schistosomiasis and has begun to be used on a large scale in mass drug-administration programmes [40]–[43] . Microscopical parasitology , particularly the Kato-Katz method , has generally been used to estimate epidemiological parameters prior to commencement of these control programmes . If , however , control is successful the rate of schistosome egg excretion will decline in treated populations and Kato-Katz is likely to become ever less useful due to its relative insensitivity [3] , [11] , [17] , [34] . In the absence of effective alternative methods to detect schistosome infections directly it is now being anticipated that antibody detection methods will become increasingly useful [17] . The importance of developing novel diagnostic tools appropriate to the changing requirements of control programmes is now becoming widely recognised , despite this research area often being thought of as less important than that of others , such as drug and vaccine development [11] , [43] , [44] . Stothard has recently evaluated the potential of diagnostic tests in improving control of schistosomiasis [45] and SmSEA in the ELISA format has been shown to hold promise as a diagnostic method for the monitoring of this disease in Zanzibar [46] . SEA-ELISA tests are the frequent method of choice in travellers' medicine clinics [18] and there is also a commercially-produced SmSEA-ELISA kit now available for use in the field [46] . Indeed , SEA-ELISA formats have previously given meaningful results in endemic areas [21] , [47] . The production of egg antigens is however a fairly laborious and expensive process as the eggs have to be produced from infected animals . The use of SmCTF as a replacement for SmSEA would be a cost-effective option as SmCTF is harvested from infected snails , which are relatively easy to culture and maintain . This would also reduce the number of mice that are required , in comparison to those needed for the production of SmSEA . We estimate the overall animal-usage costs of producing SmCTF to be as much as 90% less than to produce an equivalent antigenic biomass from eggs . As well as the expense of SmSEA , use of ELISA is also problematic outside the laboratory because of power requirements for blood centrifugation and electronic reading of ODs . Point-of-care ( POC ) or rapid diagnostic tests ( RDTs ) that are scalable and cost-effective for use in the developing world are becoming increasingly useful for the diagnosis of helminth infections [11] . RDTs that work on whole blood are likely to be useful for the implementation of control programmes in countries endemic for schistosomiasis , provided their application improves the efficiency of allocating praziquantel treatments and their costs can thus be off-set at least in part by a reduction in drug-wastage as a result of those who are antibody-negative being left untreated [45] . As the SmCTF antigen is cheap to produce and appears to be as good as SmSEA in these preliminary experiments and others [27] , [28] , a rapid test incorporating SmCTF to detect anti-schistosome antibodies , in a format usable with whole blood as well as serum , is being developed . It is hoped that it can be produced for sale at a cost of <US$1 . 00 which would make its use for control of schistosomiasis in endemic areas economically justified [45] . Of course such an assay would need to be evaluated in endemic areas prior to implementation . There is currently a commercially available RDT on the market for diagnosis of both intestinal and urinary schistosomiasis by detection of circulating antigens ( marketed by Rapid Medical Diagnostics , South Africa ) , but this test has recently been shown to be unsatisfactory for diagnosis of S . haematobium infections in Zanzibar [46] , [48] , despite being a reformulation of a test that encountered problems previously . The circulating cathodic antigen ( CCA ) urine-dipstick was however found to be an effective means of testing for intestinal schistosomiasis in shoreline communities of Lake Victoria [49] . Because the CCA-RDT has been marketed at US$2 . 3–2 . 8 per test [49] , and more recently at $1 . 98 [50] , an RDT using SmCTF to detect anti-schistosome antibodies would be a more cost-effective alternative , with great benefits to endemic areas , as many are resource-poor . As the results above indicate SmCTF is as good as SmSEA in detecting anti-schistosome antibodies in ELISA and thus decidedly a promising technique for the diagnosis and monitoring of schistosomiasis . Our results also show that SmCTF reacts well with both S . mansoni and S . haematobium infection sera , an advantage over the urine-CCA strips currently on the market . Definitive diagnosis of schistosomiasis japonica is still very reliant on the demonstration of viable ova in faeces or other histological samples , despite the fact that parasitological techniques have now become relatively insensitive following widespread , repeated chemotherapy [51] . Immunodiagnostic technology was incorporated into the national control program in China in the 1980s as a way of improving identification of individuals as targets for treatment [52] . The Chinese national control program recommends the use of immunodiagnostic assays for the screening of populations in schistosome-endemic areas with a prevalence of less than 20% [53] . Immunodiagnosis is also used for preliminary screening in areas where the prevalence rate is less than 5% [52] . Among the techniques that have been successfully applied in the field is the ELISA incorporating S . japonicum soluble egg antigens ( SjSEA ) for detection of the anti-schistosome antibodies . The assay has high sensitivity and good specificity in diagnosis of S . japonicum infections [53]–[55] , but the same problems as those with the SmSEA-ELISA are encountered: SjSEA is expensive to produce , there is the problem of requiring sera for ELISA and the ELISA format is generally not suitable for use outside of a designated laboratory . The results here show that SmCTF performs equivalently to SjSEA in ELISA with S . japonicum infection sera . This suggests not only that SmCTF could replace SjSEA in ELISA for diagnosis of S . japonicum infections , but also that an RDT incorporating SmCTF for diagnosis of S . mansoni and S . haematobium infections could have potential in the diagnosis of schistosomiasis japonica . In conclusion , the SmCTF antigen appears to perform equivalently to SmSEA in ELISA with S . mansoni and S . haematobium infection sera , as well as with S . japonicum sera . SmCTF is more easily and cheaply produced than SmSEA , with the added advantage of a reduced number of laboratory animals required for antigen production . As well as replacing SmSEA in ELISA , the SmCTF could have potential in the development of an RDT that detects anti-schistosome antibodies in whole blood-usable format . If this RDT could be marketed at <US$1 . 00 per test its application in schistosomiasis control programmes in endemic areas would seemingly be justified . Development and evaluation of a RDT incorporating SmCTF as the antibody target is now underway .
|
Diagnosis of schistosomiasis is problematic since no method is yet available that gives both 100% sensitivity and 100% specificity . The method traditionally used is microscopy , but because of inherent insensitivity this technique often wrongly diagnoses patients as uninfected . Use of serological assays involving detection of specific antibodies is now increasing since the putative sensitivity of these tests is much higher than that of other alternative methods of diagnosis . They are routinely used in travellers' medicine clinics where often only light infections are encountered which microscopy is not sensitive enough to detect . ELISA incorporating schistosome soluble egg antigens ( SEA ) is often the antibody-detection test of choice . The use of the SEA-ELISA for diagnosis of schistosomiasis in developing countries is however restricted since SEA is relatively expensive to produce . Here we investigated whether a cheaper alternative S . mansoni antigenic preparation derived from schistosome cercariae ( SmCTF ) could potentially replace SEA in ELISA . Our results demonstrate that SmCTF performs equivalently to S . mansoni SEA for the diagnosis of both S . mansoni and S . haematobium infections , and that SmCTF is also as good as S . japonicum SEA for the diagnosis of schistosomiasis japonica . We discuss how even more affordable and practical diagnostic aids for schistosomiasis might be developed .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"clinical",
"laboratory",
"sciences",
"immunology",
"microbiology",
"parasitic",
"diseases",
"parasitology",
"test",
"evaluation",
"neglected",
"tropical",
"diseases",
"infectious",
"disease",
"control",
"infectious",
"diseases",
"travel-associated",
"diseases",
"biology",
"schistosomiasis",
"diagnostic",
"medicine",
"antigen",
"processing",
"and",
"recognition",
"clinical",
"immunology"
] |
2012
|
Comparison of Schistosoma mansoni Soluble Cercarial Antigens and Soluble Egg Antigens for Serodiagnosing Schistosome Infections
|
Molecular xenomonitoring ( MX ) —pathogen detection in the mosquito rather than human—is a promising tool for lymphatic filariasis ( LF ) surveillance . In the Recife Metropolitan Region ( RMR ) , the last LF focus in Brazil , Culex quinquefasciatus mosquitoes have been implicated in transmitting Wuchereria bancrofti parasites . This paper presents findings on the ideal mosquito collection method , mosquito dispersion , W . bancrofti infection in mosquitoes and W . bancrofti antigen in humans to aid MX development . Experiments occurred within two densely populated urban areas of Olinda , RMR , in July and August 2015 . U . S . Centers for Disease Control and Prevention ( CDC ) light traps were compared to battery-powered aspirators as collection methods , and mosquito dispersion was measured by mosquito mark release recapture ( MMRR ) . Female Cx . quinquefasciatus were tested by PCR for W . bancrofti infection , and study area residents were screened by rapid tests for W . bancrofti antigen . Aspirators caught 2 . 6 times more total Cx . quinquefasciatus , including 38 times more blood-fed and 5 times more gravid stages , than CDC light traps . They also collected 123 times more Aedes aegypti . Of the 9 , 644 marked mosquitoes released , only ten ( 0 . 01% ) were recaptured , nine of which were < 50m ( 34 . 8m median , 85 . 4m maximum ) from the release point . Of 9 , 169 unmarked mosquitoes captured in the MMR , 38 . 3% were unfed , 48 . 8% blood-fed , 5 . 5% semi-gravid , and 7 . 3% gravid . PCR on 182 pools ( 1 , 556 mosquitoes ) found no evidence of W . bancrofti infection in Cx . quinquefasciatus . Rapid tests on 110 of 111 eligible residents were all negative for W . bancrofti antigen . Aspirators were more effective than CDC light traps at capturing Ae . aegypti and all but unfed stages of Cx . quinquefasciatus . Female Cx . quinquefasciatus traveled short ( < 86m ) distances in this urban area . Lack of evidence for W . bancrofti infection in mosquitoes and antigen in humans in these fine-scale studies does not indicate that LF transmission has ceased in the RMR . A MX surveillance system should consider vector-specific collection methods , mosquito dispersion , and spatial scale but also local context , environmental factors such as sanitation , and host factors such as infection prevalence and treatment history .
Lymphatic filariasis ( LF ) is a neglected tropical disease and ranked by the World Health Organization ( WHO ) as the world’s leading cause of physical disability , the second leading cause of long-term disability overall , and the leading cause of disability due to infectious disease [1 , 2] . In 2000 , an estimated 120 million people were infected with LF parasites and 1 . 3 billion were considered at risk [3 , 4] . The nematode Wuchereria bancrofti is responsible for nearly 90% of global LF infections [5] . The mosquito Culex quinquefasciatus is the most common vector of urban , nocturnally periodic W . bancrofti and is thought to be the sole vector of LF in Brazil [6] . In 2000 , the Global Program to Eliminate Lymphatic Filariasis ( GPELF ) aimed to eliminate LF by the year 2020 by interrupting transmission via mass drug administration ( MDA ) and integrated vector management ( IVM ) [7] . Since then , 10 GPELF member countries have eliminated LF; of the 73 remaining , 25% are in the ‘surveillance’ phase and 60% have commenced MDA and IVM [8] . LF was introduced to Brazil in the 19th century , and in 1952 the first national LF survey found transmission in 11 states [9 , 10] . By the 1990s , after sustained control efforts throughout Brazil , LF remained in three cities in three states: Belém ( Pará ) , Maceió ( Alagoas ) and Recife ( Pernambuco ) [9] . In 2018 , only Recife and its surroundings remain a focus [9] . The Recife Metropolitan Region ( RMR ) has a population of over 3 . 7 million people in 15 municipalities , including Recife proper and the neighboring city of Olinda [11] . In the RMR , nearly 25% of residents live in favelas ( slums ) and areas of suboptimal municipal infrastructure , including proximate to many polluted water bodies ( e . g . , canals with open sewage ) that can serve as Cx . quinquefasciatus breeding sites [12] . In 2000 , overall LF prevalence by thick blood smear ( TBS ) was 1 . 34% in Recife and Olinda [13–15] . Recife began MDA in 2005 and Olinda in 2006 , with the highest priority areas receiving 5–6 rounds of MDA through 2012 . Both cities have assessed transmission by surveying children aged 6–7 years with immunochromatographic card tests ( ICTs ) to detect circulating filarial antigen ( CFA ) [16] . Molecular xenomonitoring ( MX ) is the use of molecular methods , such as polymerase chain reaction ( PCR ) , to detect pathogen DNA or RNA in the vector as a proxy for infection in the human population . MX is a promising method for monitoring LF transmission , MDA and IVM success , and LF elimination [17–21] . Over nearly twenty years , MX has been tested in a variety of LF-endemic settings with different vector-parasite dynamics , with evidence of its utility from five out of six WHO Regions: Africa/AMRO ( Ghana , Sierra Leone , Tanzania ) , Americas/AMRO/PAHO ( Trinidad and Tobago ) , Eastern Mediterranean/EMRO ( Egypt ) , Southeast Asia/SEARO ( India , Sri Lanka ) , and Western Pacific/WPRO ( American Samoa , French Polynesia , Samoa ) [21–30] . MX could prove more appropriate and useful as control activities reduce LF transmission , because after MDA parasitological detection tools such as TBS become less sensitive while immunological detection tools such as ICTs become less specific [31] . Moreover , it would be necessary to screen large population samples to detect low and clustered transmission areas . For MX , it is crucial to capture adult female mosquitoes so that they can be screened for infection ( any parasitic stage ) as well as infectivity ( the L3 larval stage ) , the latter being the most precise determinant of transmission potential . Several MX protocols have been developed for Cx . quinquefasciatus [21 , 32–35] . Currently , there is no universally recommended strategy for MX sampling or tool for MX collection and MX programs differ by site-specific vector and parasite dynamics . In Tanzania , for example , U . S . Centers for Disease Control and Prevention ( CDC ) gravid traps collected the greatest number of Cx . quinquefasciatus of all stages as well as gravid stages in relation to four methods . However , a subsequent comparison of CDC light vs . CDC gravid traps demonstrated that both caught similar numbers of mosquitoes , although of different gonotrophic status , and that CDC light traps collected more infected mosquitoes [24] . In the RMR , the preferred method for adult mosquito collection is aspiration , which also has the benefit of preferentially collecting post-blood meal , resting mosquitoes . This is advantageous for LF elimination and pathogen surveillance because blood-fed , gravid , and semi-gravid mosquitoes are more likely to have ingested mf-infected blood . To date , a collection method comparison ( CMC ) for Cx . quinquefasciatus including aspiration has not been published from any urban setting . Therefore , there is a dearth of evidence on collection tools , such as aspirators , and program-oriented techniques for Cx . quinquefasciatus mosquitoes , which dominate in urban areas [19] . For instance , in the RMR , fixed battery-operated or energy-source requiring traps ( e . g . , BG sentinel , CDC light ) are discouraged due to battery theft and power cuts , whereas trapping methods that rely on attractants ( e . g . , gravid , sticky ovitraps ) cannot be placed inside domestic spaces due to residents’ distaste for the strong odors emitted . A CMC including aspiration in this setting would aid LF elimination and MX system development planning . Ideally , MX methodology should take into account the geographical scale and directionality of mosquito dispersion as the former is related to the spatial scale of disease transmission [36 , 37] . Understanding mosquito dispersion in a given setting allows public health officials to more accurately plan the limits of where related vector borne disease may occur , and thus where control efforts should concentrate . Although some studies have included evaluations of different mosquito collection methods , none have formally assessed mosquito flight distance and patterns within the context of MX and none has occurred in a densely populated , urban area [21 , 24 , 28 , 30 , 38–40] . One of the most straightforward methods for measuring mosquito dispersion is mosquito mark-release-recapture ( MMRR ) , but most MMRR studies of Cx . quinquefasciatus have been conducted in high-income countries ( e . g . , United States ) and among rural settings ( e . g . , dairy farms ) [36 , 41–44] . While rural MMRR studies indicate that Cx . quinquefasciatus can travel up to 2 km for host blood seeking and oviposition , the only published ‘urban’ Cx . quinquefasciatus MMRR study occurred in a central Texas university town of approximately one eighth the population density of RMR favelas [45] . For comparison , studies on Ae . aegypti dispersal , including in urban areas of Brazil ( often set within less population dense / forested areas of cities ) , indicate that Ae . aegypti tend to fly 100m or less [46 , 47] . Despite the promise of MX for LF and other vector-borne pathogens , there are no published reports of its use to detect W . bancrofti infection in Cx . quinquefasciatus in urban settings . CMC studies could provide information on effective , practical , and acceptable collection tools for MX programs . MMRR studies could provide crucial insights on LF risk and transmission , especially if mosquito parameters ( e . g . , mean distance travelled , MDT ) are combined with those on human infection over the same space and , ideally , time . As LF elimination efforts continue , and eventually are localized to difficult-to-treat urban areas , information on mosquito dispersion in such settings is of increasing importance . The following strategies were employed to develop a MX system in the RMR: ( i ) CMC to determine whether battery-powered aspirators or CDC light traps more efficiently collect Cx . quinquefasciatus females; ( ii ) MMRR to estimate mosquito dispersion to determine grid size for use in subsequent surveillance; ( iii ) molecular screening via PCR in a sample of female Cx . quinquefasciatus to determine W . bancrofti infection in mosquitoes; and ( iv ) immunological screening via ICT to detect W . bancrofti antigen in study area residents .
The CMC , MMRR , molecular screening of mosquitoes , and immunological screening of study residents were conducted in two selected areas within the neighborhood of Sítio Novo in the city of Olinda , RMR . Olinda is the second most populous and population-dense city of the RMR , with 377 , 779 residents in its area of 41 . 68 km2 ( Fig 1 ) [11] . It has a tropical monsoon climate ( Köppen climate classification = As ) , and temperatures range from 30 °C ( 86 °F ) in January and February to 21 °C ( 70 °F ) in July [48] . Peak dry season is in November ( average 36mm rainfall ) while the rainy season , extending June—August , peaks in July ( average 388 mm rainfall ) [49 , 50] . Data collection occurred between July 22 and August 21 , 2015 , coinciding with the end of rainy season and associated peak in mosquito abundance . Houses were selected using satellite images and geographic information systems ( GIS ) software of ArcGIS 10 . 2 ( ESRI 2014 . ArcGIS Desktop: Release 10 . Redlands , CA: Environmental Systems Research Institute ) and QGIS 2 . 10 . 1 ( QGIS Development Team ( 2015 ) . QGIS Geographic Information System . Open Source Geospatial Foundation Project . http://qgis . org ) . House selection accounted for geographic ( aligning along transport arteries in CMC ) and environmental ( e . g . , avoiding mangrove in MMRR ) barriers , as well as local health authority advice on the most secure areas to work . In the field , study teams used a combination of global positioning system ( GPS ) devices ( Garmin GPSmap 76cs , 3m precision ) and GIS / satellite image maps to locate selected houses . If residents were not willing or able to participate , including providing regular access over four weeks , then alternative houses were enrolled by selecting houses to the right , then left , then opposite the initial house until an appropriate alternative could be found . To avoid contamination , the CMC and MMRR study areas were separated by a buffer zone of 100m based on the estimated average mosquito flight distance from urban Ae . aegypti dispersion studies ( Fig 1 ) . The CMC occurred in a commercial and residential zone with some paved streets , municipal sanitation , and drainage systems . Houses were of higher quality construction , with brick walls , solid/permanent roof , some partially screened windows , and fewer wall openings , than those in the MMRR area . Still , much of this area was considered to be of suboptimal housing , including favelas . As much as possible , houses were selected along main streets in order to provide better access for equipment transfer ( Fig 2 ) . The MMRR occurred in an infrastructure-lacking residential area with poorly paved streets , sanitation , and drainage . During the study period , houses were often flooded from an adjacent area of riverine mangrove ( Fig 3 ) . Molecular screening of Cx . quinquefasciatus and immunological screening of study participants occurred in CMC and MMRR study areas , from where the samples for each were obtained . Mosquito collection nets were placed in an open-top storage box and transported back to the IAM/FIOCRUZ Insectary within two hours of field collection . Upon arrival , nets were immediately placed in a -20°C freezer for at least 20 minutes to immobilize the mosquitoes . Mosquitoes were then removed from the freezer and placed on ice for identification , sex determination , and assessment of female physiological status . The numbers and status of female Cx . quinquefasciatus mosquitoes were recorded per house , per day . Female Cx . quinquefasciatus and Ae . aegypti mosquitoes were placed in Eppendorf tubes ( maximum of 50 per tube , separated by species ) labeled per house per day and stored in a -80°C freezer for future analysis . Meteorological data ( temperature , humidity , wind ) that could influence mosquito flight range , survival and dispersal were obtained from the Brazilian National Meteorological Institute ( INMET: www . inmet . gov . br ) and the Pernambuco State Agency for Water and Climate ( APAC: www . apac . pe . gov . br ) [49 , 50] . Female Cx . quinquefasciatus mosquitoes were pooled into groups of up to 10 per pool depending on study area . Pooling was done by house per day ( MMRR ) or by house per week ( CMC ) . RNA was extracted using a Ambion Trizol-based protocol ( see appendix III ) and RNA was re-suspended in 30 μl of Invitrogen Ultrapure water and stored at -80°C to preserve RNA prior to reverse transcription . RNA samples were reverse transcribed using a QIAGEN QuantiTect reverse transcription kit according to manufacturer’s instructions . Successful generation of cDNA was confirmed by real time PCR assays targeting the Cx . quinquefasciatus S7 mRNA gene using QIAGEN QuantiTect Sybr Green Master Mix . W . bancrofti detection was undertaken using a Taqman real time PCR assay targeting the constitutively expressed tph-1 gene using Promega GoTaq Probe qPCR Master Mix [54] . See Appendix III for more details . Concurrent to CMC and MMRR activities , immunochromatographic card tests ( AD12-ICT card test , NOW Filariasis ) were requested from all eligible ( age 2–65 years ) and consenting residents of the 35 houses in this study . This test detects CFA using the monoclonal antibody AD12 , which recognizes a 200-kDa filarial antigen from either adult worms or microfilariae . [55] Study teams approached each HoH and any available household members , presented an information sheet and an invitation to receive ICT screening at the local community center . Any resident who did not attend the community center was visited at least three times to attempt to administer the ICT in their residence . The test was performed according to the manufacturer’s instructions and read by trained technicians in the field after 10 minutes . Visualization of two lines ( test and control ) was interpreted as a positive result . Data were double entered by two independent data entry staff , cleaned , and analyzed with Stata 14 ( StataCorp . 2015 . Stata Statistical Software: Release 14 . College Station , TX: StataCorp LP ) . Study aims and methods were presented to HoHs and verbal and written informed consent was sought; households were enrolled upon receipt of written informed consent . All names , addresses , and GPS coordinates of participating HoHs and residents were concealed from study staff apart from the principal investigator and study coordinator , both of whom held the linking keys . Field teams worked during the mornings of weekdays due to security concerns as well as to increase acceptability of daily aspiration or CDC light trap placement/net collection . Ethical approval was obtained from the Research Ethics Committees of the Instituto Aggeu Magalhães ( IAM/FIOCRUZ ) and the London School of Hygiene & Tropical Medicine ( LSHTM ) prior to the commencement of fieldwork . [CAAE: 44535515 . 0 . 0000 . 5190; LSHTM: 10276; 10185] .
Of a total of 80 trapping nights planned , 78 were obtained for battery-powered aspiration , and 68 for CDC light traps . The primary reasons for losses in trapping night measurements were battery failure ( especially for CDC light traps left overnight ) and inability to enter participants’ houses . Table 1 presents the number of mosquitoes collected by collection method , species and physiological status . A total of 970 adult females of Cx . quinquefasciatus ( unfed = 393 , blood-fed = 403 , semi-gravid = 165 , gravid = 9 ) were captured , of which 684 were by aspiration and 286 by CDC light traps . A total of 188 Ae . aegypti were captured , all but one by aspiration . Adjusting for the house and night factors , aspirators caught 2 . 6 times more total females , and 38 times more blood-fed mosquitoes than CDC light traps ( all p<0 . 0001 ) . Aspirators caught 5 . 8 times more gravid and semi-gravid mosquitoes than CDC light traps; these abdominal conditions were pooled due to the small number ( nine ) of semi-gravid mosquitoes collected . Aspirators collected almost 25% fewer unfed ( p<0 . 0001 ) Cx . quinquefasciatus than CDC light traps . Aspirators collected 123 times more ( p< 0 . 0001 ) total females of Ae . aegypti . Data from the experimental period ( R2 –R4 ) were collected over a period of 19 days in 25 households , for a total number of 475 observations . The study recruited 24 houses as planned , but one house refused access to the field team after the first week , so another was recruited in its place to preserve measurements across a theoretical grid ( the size of which would be measured along radii emanating from the CRP ) . This newly recruited 25th house then participated for 3 weeks , yielding four weeks of collections from each of 23 houses , one week from the first house that dropped out , and the remaining three weeks from the 25th house . In total , the CMC and MMRR experiments collected 10 , 139 ( 970 CMC , 9169 MMRR ) Cx . quinquefasciatus and 910 ( 188 CMC , 722 MMRR ) Ae . aegypti female mosquitoes ( Table 3 ) . All Ae . aegypti and Cx . quinquefasciatus mosquitoes were stored at -80C to preserve RNA ( filarial , arboviral ) for future analysis; a subset ( 15% of the total yield ) of Cx . quinquefasciatus was then subjected to molecular analysis . Of the 10 , 139 Cx . quinquefasciatus collected ( Table 3 ) , 182 pools ( 112 CMC , 70 MMRR ) representing 1 , 556 ( 856 CMC , 700 MMRR ) female mosquitoes of all abdominal conditions were screened for W . bancrofti infection . PCR analysis confirmed successful generation of Cx . quinquefasciatus cDNA from each mosquito pool but revealed no evidence of W . bancrofti infection . A total of 110 ( 99 . 1% ) out of a reported 111 full and part-time residents of the 35 houses included in the CMC and MMRR studies underwent immunological analysis via ICT . The majority were female ( 63% ) , and the gender disparity was more evident in older age groups . Nearly 25% of the population undergoing ICT was 61 years of age or older . None tested positive for W . bancrofti CFA .
This study compared battery-powered handheld aspirators with CDC light traps , although several other methods had been considered . Gravid traps were rejected due to acceptability concerns related to the smell of attractants ( e . g . , grass infusion ) if used indoors , logistical issues related to transporting large volumes of infusions , and trap placement in relation to security ( e . g . theft ) given extremely limited secure outdoor space for each house in the study . A paper by Irish et . al . found that the gravid traps caught less infected Cx . quinquefasciatus mosquitoes than CDC light traps [24] . BG sentinel traps were rejected due to their large size , unwieldiness and fan noise . Sticky ovitraps were rejected based on local expert advice and experience that they are vastly inferior to battery-powered aspiration , and genetic material ( RNA ) in collected mosquitoes would likely be degraded due to desiccation . The nearly three-fold superiority of aspiration in collecting female Cx . quinquefasciatus may be surprising , given that many other sites preferentially use CDC light or gravid traps for this species [24 , 26 , 27 , 38 , 40 , 57 , 58] . However , much of the existing literature is based upon studies conducted in rural settings with different residential and sanitation infrastructure and low population density . One previous study in the metropolitan area of Recife found that CDC light traps collected an average of 55 Cx . quinquefasciatus females per trap in 1991–2 [59] . While this quantity is much greater than that found in the current study , one possible explanation of this result is an improvement of infrastructure and sanitation within Olinda over the past two decades . Aspirators collected 25% fewer unfed Cx . quinquefasciatus than CDC light traps , consistent with CDC light traps preferentially attracting pre-blood meal and aspirators collecting post-blood meal mosquitoes [24] . Furthermore , aspirators collect mosquitoes resting indoors , which are less likely to be unfed females [60] . As female Cx . quinquefasciatus mosquitoes are endophagic and endophilic , battery-powered aspiration inside houses should be more likely to collect resting blood-fed females . This was the case in the Sítio Novo , where aspirators collected 47 times more blood-fed Cx . quinquefasciatus than CDC light traps . CDC light traps collected less than 2% of the blood-fed females , which is much lower than most previous studies [30 , 40 , 60] , although in line with one recent study in Tanzania [24] . Several other limitations should be noted . First , there were more operational issues surrounding the use of CDC light traps than aspirators . Light and noise emitted from the traps were aggravating to several residents; three participating households requested the CDC light traps be removed from their bedrooms at night . Of trapping nights lost , nine were due to battery failure , two were lost due to traps being prematurely switched off , and one was due to participants not being at home . By contrast , only two data points were lost during aspiration , both due to participants not being home during morning visits . This also raised another important issue . As CDC light traps require a freshly charged battery each night , if a house cannot be accessed each morning , then the previous night’s collection net cannot be retrieved and a new battery cannot be swapped . This effectively means that not being able to access a house during CDC light trap testing results in losing two nights’ of trapping data , whereas not being able to access a house for aspiration results in only one data point being lost . Of the 188 Ae . aegypti captured , 99% were collected by aspirators; so , unlike CDC light traps , they may also offer an alternative tool to sticky ovitraps for collecting adult female Ae . aegypti [30 , 40 , 61–67] . The finding that aspirators collect adult Ae . aegypti extremely well , and the possible co-circulation of arboviruses with LF , indicate that a combined MX and surveillance program for several vector-borne diseases could be both time- and cost-effective [68] . Aspiration of resting mosquitoes is not a new collection method for vectors of LF , and has already been successfully adopted for xenomonitoring surveillance during and after MDA programs in Egypt and India [69 , 70] . However , normally these studies involved other collection methods ( e . g . , CDC light traps ) , and not the large , battery-powered aspirators used in the current study . The type of aspirator used here provides a promising tool for a xenomonitoring program for the RMR . While aspiration has for some time been the locally preferred method of collecting adult resting mosquitoes by Secretary of Health officials , no standard operating procedures have previously been in place . The present study produced an easy to use written protocol that local researchers ( including those in other research groups ) are currently using in order to standardize adult vector collection . Since aspiration tends to collect a significantly higher proportion of blood-fed mosquitoes than some other methods , any PCR-positive samples could , in principle , be attributed to the mosquito having recently ingested an infected blood meal , as opposed to carrying an established infection . In an MX program , this could potentially inflate the infection rate beyond the true transmission potential [71] . It has likewise been argued that other collection methods that preferentially capture older and previously blood-fed mosquitoes , such as gravid traps , would have a higher likelihood of detecting infective L3 larvae [38] . The introduction of a reverse transcriptase based PCR assay however , would negate the need to exclude blood-fed mosquitoes , as its mRNA based primers are designed to detect L3 specific larvae , so could therefore give an estimation on vector infectivity rates and a direct measure of transmission potential [54] . This study was conducted to determine the flight range , survival and dispersal of adult Cx . quinquefasciatus and hence set spatial resolution in a gridded MX system . In this densely populated urban area , the median flight range was 35m , the majority ( 90% ) dispersed within 50m , and the maximum flight range was 85m from the CRP . Of 9644 ( 7235 female ) marked mosquitoes , a total of 10 ( 8 female , 2 male ) , or 0 . 103% , were recaptured . Although this is shorter than other distances reported in the scant literature available on the flight range of Cx . quinquefasciatus , the most likely explanation is that the richness in host/breeding site availability provided in the urban environment means that a female mosquito does not need to fly far to find blood for egg development or water for oviposition . Although results from experiments with low or zero recapture rates may be less likely to be published , a recent review of published studies indicates variation in recapture rates between zero and 14% [72] . The recapture rates found in the current study are low but are within the range of MMRR studies for Cx . quinquefasciatus , which tends to have lower recapture rates compared with other mosquito species [36] . While other methods such as sticky ovitraps could have been used , the requirement of preserving filarial RNA meant that field teams would have had to collect sticky tapes daily or more frequently due to the intense heat and potential predators in the field site; these issues rendered such tools impossible for use . Aspiration took place within houses and in the peri-domestic area , so it is possible that study teams may have missed marked mosquitoes that did not travel indoors but remained in the narrow pathways of the study site . The low recapture rate may also relate to the marking procedure and its effects . First , marked mosquitoes may have had a lower survival rate compared to wild mosquitoes . Although the effect of the marking was found to be small in the pilot experiment , mosquitoes may still have been harmed during the procedure or transport towards the field site . Second , the color of the mosquitoes removes advantages of their natural camouflage and is likely to make them more vulnerable to predation . Additionally , it is possible that recapture rates may have been higher if collection methods not utilized in the present study , e . g . , BG Sentinel traps , were used . In this study , all mosquitoes were recaptured within a period of four days . This is in concordance with the literature , where recapture success decreases after approximately four to six days [42 , 52 , 73] . Anecdotally , members of the community reported seeing or killing colored mosquitoes . In particular , the owner of the house on whose property the CRP was located repeatedly reported seeing colored mosquitoes inside the house up to five days after release but these mosquitoes were not recaptured by study teams . One household member of a participating house accidentally killed a magenta-colored mosquito , saved it , and presented it to study teams as evidence ( and with an apology ) : it was a blood fed female mosquito from R3 that had travelled 23 . 7m before it was killed , reportedly 10–12 hours after its release . Fewer Ae . aegypti were collected than Cx . quinquefasciatus , which is unsurprising given the collection method and deployment schedule . As this study was conducted to design a MX system for LF , the MMRR was primarily interested in Cx . quinquefasciatus flight distance and survival . Battery-powered aspirators were chosen because they preferentially collect post-blood meal resting females and aspiration occurred from 9am– 11:30am each day in order to coincide with resting Cx . quinquefasciatus . Adjusting the aspiration schedule towards later in the day would have likely resulted in collecting more Ae . aegypti mosquitoes . The furthest recorded distance travelled was 85 . 4m from the CRP , on day three and at the outer limit of the study area . Hence mosquitoes may have also dispersed beyond the study area . This is difficult to confirm , although Cx . quinquefasciatus has been reported to travel over 15km ( 810–1680m ) from a release site in other studies [36 , 41 , 43 , 53 , 74 , 75] . However , the need to migrate long distances seems relatively low in this study area , given the availability of human blood meals in the densely populated urban setting . On the other hand , the relatively small size of the study area may have biased the observed MDT downwards [36] . Although the sample size of recaptured marked mosquitoes was insufficient to perform statistical analysis , most mosquitoes appeared to travel upwind , despite the relatively high wind speeds recorded over the study period . Reisen et al . suggested that Cx . quinquefasciatus may travel towards areas with vegetation to seek protection against the wind , although Schreiber et al . found they dispersed mainly downwind regardless of land cover [41 , 73] . This study required high participation rates from the community , requesting access to every room in participants’ house on a daily basis over four weeks . One house refused access to the field team after the first week and access to other houses was denied on an incidental basis . Reported reasons for refusal were inconvenience caused by the procedure , having visitors and the conception that mosquitoes would return the next day . This study aimed to develop a MX system for the RMR , with the primary interest being in ideal collection method and ascertaining the limits of mosquito dispersion . Unfortunately , the expense of field collection resulted in a limited ability to conduct molecular screening of mosquito samples for W . bancrofti infection in this current study . Thus , researchers decided to screen a proportion of mosquitoes from each area , and biobank the rest with the hopes of securing future funding for further analysis . The absence of W . bancrofti infection in Cx . quinquefasciatus mosquitoes and the absence of antigen against W . bancrofti in humans in this small study area does not prove the absence of LF transmission in the RMR . As current infection rates in the active foci are estimated to range between 0 . 6 and 2% , due to repeated rounds of MDA , much larger sample sizes ( >20 , 000 mosquitoes ) would be required to detect W . bancrofti [9 , 28] . Mosquitoes collected for MX of LF are potentially useful for monitoring of other infections , in particular arboviruses such as dengue and Zika , although only if the necessary storage and processing protocols are followed to prevent RNA degradation . Even ignoring the likely clustering of infection , the upper 95% confidence limit for the zero positive tests out of 110 is a prevalence of 3 . 4% [76] . Among over 10 studies reporting MX program results , the majority have originated from the African ( AMRO ) , Southeast Asian ( SEARO ) , and Eastern Mediterranean ( EMRO ) Regions of the WHO . This is only the second study to report MX results , however preliminary , from the Pan American ( PAHO ) region . Moreover , only one MX program has evaluated aspirators in collecting Cx . quinquefasciatus for W . bancrofti detection . In contrast to several other studies , this study found overwhelming evidence that large , handheld battery-powered aspirators are extremely effective tools for collection of adult Ae . aegypti as well as Cx . quinquefasciatus irrespective of physiological status with the exception of unfed females [60 , 77–80] . It should be emphasized that the handheld aspirators used in this study are significantly larger than the handheld , backpack , or mouth aspirators that have been used in other sites ( Appendix II ) . This research identifies a role for battery-powered aspiration for MX , having demonstrated that they are extremely effective for collecting not only Cx . quinquefasciatus but also Ae . aegypti adult females in this densely populated urban setting . This demonstrates that MX may be promising and feasible where there is the possibility of an integrated LF and arbovirus surveillance program . Although few in number , the recaptured mosquitoes suggest a suitable grid size for MX sampling may be 75 x 75m or slightly larger , based on 90% of mosquitoes dispersing at least 50m and at least one up to 85m . It is possible that in less densely populated or built up urban areas a slightly larger grid ( e . g . , 100m x 100m ) may suffice . This research team recommends prioritizing considerations of spatial scale and transmission dynamics , including underlying human infection prevalence , when designing grid-based MX systems . MX may seem to require substantial up-front investment in monitoring mosquito populations , especially when human health data ( e . g . , physician confirmed disease or lab confirmed infection ) may already be available . However , with correct design and sufficient time for deployment , mosquito-based MX has the potential to enhance current LF surveillance systems ( as well as potentially aid in the early warning of new and cyclical infections such as arboviruses ) . In settings like the RMR—where , in addition to LF , microcephaly , Zika virus , dengue virus , and chikungunya virus have caused enormous strain on public health resources in recent years—such enhanced disease surveillance systems could be very helpful for planning the allocation of public health resources .
|
Lymphatic filariasis ( LF ) is a parasitic disease transmitted by mosquitoes , and can cause elephantiasis . It is the world’s leading cause of disability due to infectious diseases , affects over 120 million people globally , and is scheduled for global elimination via mass drug administration ( MDA ) and mosquito control . Molecular xenomonitoring ( MX ) is a process of screening mosquitoes—not humans—for parasites to estimate whether they are circulating in human populations . MX is especially useful during and following MDA , when new case detection becomes difficult , but is challenging to design and conduct in cities . Using two study sites in the Recife Metropolitan Region , Brazil , we investigated two crucial questions for urban MX development—“What is the best operationally feasible tool to catch adult mosquitoes ? ” and “How far do mosquitoes disperse in cities ? ”—in order to determine placement of future surveillance sites . We also screened a proportion of mosquitoes and all eligible residents from the study sites for LF infection . We determined that handheld battery powered aspirators were the best mosquito collection tool; that mosquitoes flew no more than about 85m; and—in this small sample of mosquitoes and very small sample of humans—there was no evidence of LF infection in mosquitoes or study area residents .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"geographical",
"locations",
"tropical",
"diseases",
"vector-borne",
"diseases",
"parasitic",
"diseases",
"animals",
"filariasis",
"neglected",
"tropical",
"diseases",
"insect",
"vectors",
"lymphatic",
"filariasis",
"infectious",
"diseases",
"geography",
"wuchereria",
"bancrofti",
"south",
"america",
"wuchereria",
"culex",
"quinquefasciatus",
"disease",
"vectors",
"insects",
"brazil",
"arthropoda",
"people",
"and",
"places",
"helminth",
"infections",
"mosquitoes",
"eukaryota",
"urban",
"areas",
"earth",
"sciences",
"geographic",
"areas",
"nematoda",
"biology",
"and",
"life",
"sciences",
"species",
"interactions",
"organisms"
] |
2018
|
Development of an urban molecular xenomonitoring system for lymphatic filariasis in the Recife Metropolitan Region, Brazil
|
Variation in susceptibility to infectious disease often has a substantial genetic component in animal and plant populations . We have used genome-wide association studies ( GWAS ) in Drosophila melanogaster to identify the genetic basis of variation in susceptibility to viral infection . We found that there is substantially more genetic variation in susceptibility to two viruses that naturally infect D . melanogaster ( DCV and DMelSV ) than to two viruses isolated from other insects ( FHV and DAffSV ) . Furthermore , this increased variation is caused by a small number of common polymorphisms that have a major effect on resistance and can individually explain up to 47% of the heritability in disease susceptibility . For two of these polymorphisms , it has previously been shown that they have been driven to a high frequency by natural selection . An advantage of GWAS in Drosophila is that the results can be confirmed experimentally . We verified that a gene called pastrel—which was previously not known to have an antiviral function—is associated with DCV-resistance by knocking down its expression by RNAi . Our data suggest that selection for resistance to infectious disease can increase genetic variation by increasing the frequency of major-effect alleles , and this has resulted in a simple genetic basis to variation in virus resistance .
Variation in susceptibility to infectious disease often has a substantial genetic component in animal and plant populations [1]–[4] . As pathogens are a powerful selective force in the wild , natural selection is expected to play an important role in determining the nature of this genetic variation . Selection for resistance to infectious disease can change rapidly , as new pathogens appear in the population [5] , or existing pathogens evolve , for example to evade or sabotage host defences [6] . This selection can result both in positive selection increasing the frequency of mutations that generate new resistance alleles [7] , [8] , and balancing selection stably maintaining resistant and susceptible alleles of a gene [9] . Over the last decade , genome-wide association studies ( GWAS ) have provided a more complete picture of the genetic architecture of disease susceptibility [10] . The majority of these studies have investigated non-communicable diseases in humans , and while many polymorphisms associated with susceptibility have been identified , these often have small effects and together can explain only a small proportion of the heritability [11] ( but see [12] ) . It has been suggested that the reason for this is that new mutations that increase susceptibility to non-communicable disease may tend to be deleterious , so alleles that have large effects are either removed from the population or kept at a low frequency by purifying selection [11] . However , both GWAS and classical linkage mapping studies suggest that the genetic architecture of infectious disease susceptibility may be qualitatively different [13] , as major effect polymorphisms that protect hosts against infection have been identified in many organisms , including plants [11] , humans [3] , [13] and insects [7] , [14] . Furthermore , these polymorphisms are often under strong positive or balancing selection [7]–[9] . It has therefore been argued that natural selection may cause the variation in infectious disease susceptibility to have a simpler genetic architecture than non-communicable diseases as major-effect alleles can reach a higher frequency in populations [13] . In arthropods , several studies suggest that susceptibility to infectious disease may often be affected by major-effect polymorphisms ( e . g . [7] , [14] , [15] ) . In Drosophila melanogaster , linkage mapping has been used to identify major effect resistance polymorphisms that affect susceptibility to both the sigma virus ( DMelSV ) [16] , [17] and parasitoid wasps [18] . In the case of DMelSV , two of these loci have been identified at the molecular level ( ref ( 2 ) P and CHKov1 ) , and they have been found to be common in natural populations [7] , [14] , [19] . In addition , polymorphisms in known immunity genes have been found to affect susceptibility to bacterial infection , and some of these have substantial effects [2] , [20] , [21] . To understand how natural selection affects the genetics of disease susceptibility , we have used GWAS to examine the effects of selection for resistance to pathogens on patterns of genetic variation . To do this we infected D . melanogaster both with viruses that naturally occur in this species and viruses isolated from other species . The two of the viruses that naturally infect D . melanogaster are Drosophila C Virus ( DCV ) , which is a positive sense RNA virus in the Dicistroviridae that infects a range of Drosophila species [22] , [23] , and the sigma virus DMelSV , which is a rhabdovirus that is a specialist on D . melanogaster [16] . The other two viruses naturally infect other insect species are DAffSV , which is another sigma virus that naturally infects Drosophila affinis [24] , [25] and Flock House Virus ( FHV ) , which is a nodavirus that was isolated from beetles but can infect an extremely broad range of organisms [26] . We found that the heritability of susceptibility to the two natural D . melanogaster viruses is high due to a small number of common major-effect polymorphisms . In contrast there is less genetic variation in susceptibility to viruses isolated from other species , and here there is no evidence of major effect polymorphisms .
To investigate genetic variation in resistance to viruses , we injected 47 , 220 flies from 185 different inbred lines from the Drosophila Genetic Reference Panel ( DGRP ) with four different viruses ( Table 1; note that the DMelSV data , but not this analysis , has been published before [7] ) . The extent of genetic variation in susceptibility varied considerably between the different viruses , with the greatest genetic variation being present when flies are exposed to viruses that infect D . melanogaster in nature . Comparing the two viruses where resistance was measured in terms of survival time—DCV and FHV—we found DCV resistance has significantly greater heritability ( Table 1 ) . When the two sigma viruses , DMelSV and DAffSV , are compared , again the heritability is significantly greater in resistance to the naturally occurring virus DMelSV ( Table 1 ) . While differences in heritability can be caused by differences in genetic or environmental variation , it is clear that there is genetic variation in resistance to the natural pathogens of D . melanogaster . In the case of DCV and FHV , DCV has the greater coefficient of genetic variation ( Table 1; CVg ) [27] . It is not possible to calculate the coefficient on variation for the sigma virus data as it is analysed on a logit scale . However , by inspecting Ve and Vg in Table 1 , it is clear that the differences in the heritability of resistance to DMelSV and DAffSV are primarily driven by differences in Vg . In all cases the genetic correlation in the level of resistance to different viruses is low , indicating that different genes are controlling resistance to different viruses ( Table 2 ) . In particular , the sigma viruses ( DMelSV and DAffSV ) showed no evidence of any genetic correlation , despite being relatively closely related [24] , [25] . Despite being small , there is a significant positive genetic correlation in susceptibility between three pairs of viruses , indicating that there may be some variation in the ability to survive viral infection in general . The low genetic correlations also confirm that we are measuring susceptibility to the different viruses and not an artefact of the injection procedure . To identify polymorphisms that are associated with resistance to the four viruses , we performed genome-wide association studies using the published genome sequences of the DGRP lines [28] . To correct for multiple tests and obtain a genome-wide significance threshold , we permuted the trait data across the lines and repeated the GWAS 400 times , each time recording the lowest P-value across the entire genome . Quantile-quantile ( qq ) plots of the P-values show that there are highly significant associations in the experiments using DCV and DMelSV — the two viruses that infect D . melanogaster in the wild — but not in the experiments using FHV and DAffSV ( Figure 1 ) . When the P-values are plotted along the chromosomes , it is clear that the most significant P-values cluster together ( Figure 2 , Figure S1 ) . In the case of DMelSV there is a cluster of significant SNPs around CHKov1 on chromosome arm 3R , which is a gene where we have previously shown that a transposable element insertion is associated with resistance to this virus [7] ( Figure 2 , Figure S1 ) . The second most significant cluster of SNPs in this experiment falls just below the genome-wide significance threshold , and is on chromosome arm 2L ( Figure 2 , Figure S1 ) . The SNPs in this cluster are all in strong linkage disequilibrium with a polymorphism in ref ( 2 ) P that is known to cause resistance ( the causal polymorphism was genotyped by PCR and included in this analysis ) , [8] , [14] . In the case of DCV there is a cluster of significant SNPs in and around a gene on chromosome arm 3L called pastrel , which has not previously been implicated in antiviral defence . There were no significant associations with susceptibility to FHV or DAffSV using a genome-wide significance threshold of P<0 . 05 ( Figure 2 , Figure S1 ) . We repeated the GWAS accounting for the effects of the polymorphisms in ref ( 2 ) P , CHKov1 and pastrel . The quantile-quantile plots of the resulting P-values ( Figure 1 , red points ) show that these genes can account for all of the large excess of highly significant associations with DCV and DMelSV resistance . The resulting distribution of P-values resembles that seen for the other two viruses that do not naturally infect D . melanogaster . To investigate how much of the genetic variation in susceptibility is explained by our GWAS , we calculated the proportion of the heritability that is explained by these genes . Assuming the polymorphisms have additive effects , then their contribution to additive genetic variation is 2pqa2 , where p and q are the frequencies of the alleles , and a and −a are the genotypic values of the resistant and susceptible homozygotes . In the case of DCV , pastrel ( 3L:7350895 ) can explain 47% of the heritability . In the case of DMelSV , ref ( 2 ) P explains 8% of the heritability , the doc element insertion in CHKov1 explains 29% of the heritability , and in combination these polymorphisms explain 37% of the heritability . The proportion of the heritability explained by these polymorphisms may be biased by two factors . First , we are injecting the virus , which is an unnatural route of infection , and in the case of the sigma viruses we are assaying a symptom of infection rather than viral titres or effects on host survival . Second , we can only estimate the amount of genetic variation explained in inbred lines , and this can only be directly extrapolated to outcrossed populations if all the genes affecting resistance are additive . Unfortunately , we cannot estimate the importance or direction of this bias as we only used inbred lines ( the only one of the three genes for which levels of dominance has been investigated is ref ( 2 ) P , where heterozygotes have intermediate levels of resistance when injected with the virus [29] ) . However , the bias could be substantial if we make the extreme assumptions about dominance . If the susceptible pastrel allele is recessive and only half the remaining genetic variance is additive , this polymorphism will explain 84% of Va in an outcrossed population . Conversely , if the resistant pastrel allele is fully dominant and all the remaining genetic variance is additive , this polymorphism will explain just 7% of Va in an outcrossed population . In addition to these major-effect polymorphisms , there were also other suggestive results . The most significant association for DAffSV was a synonymous SNP in scavenger receptor C1 ( 2L:4123156 A/T; individual P = 6 . 41×10−8; genome-wide permutation P = 0 . 18 ) . This gene functions both as a pattern recognition receptor of bacteria [30] and allows the uptake of dsRNA into cells [31] . A polymorphism in the gene Anaphase promoting complex 7 was associated with a 3 . 7 day increase in survival after injection with DCV ( X:6491634 G/T; individual P = 1 . 95×10−15; genome-wide permutation P<0 . 05 ) . This was at a low frequency , with the resistant variant present in 4 of 145 lines . Furthermore , it is a synonymous polymorphism , suggesting that it may not be a causal variant . The QQ plots also show that there is an excess of small P-values in three of the analyses ( Figure 1 ) , suggesting that there may be many more polymorphisms to be discovered , or that there is some unidentified population stratification . As the polymorphisms in pastrel , CHKov1 and ref ( 2 ) P have a large effect on resistance , we repeated the GWAS taking account of these polymorphisms by including them as fixed effects in the model . However , this did not lead to the identification of additional SNPs associated with resistance ( Figure S2 ) . The most significant association with DCV resistance remained Anaphase promoting complex 7 ( X:6491634 G/T ) . For DMelSV it was a SNP in the intron of off-track ( 2R:7899322 A/T , genome-wide P = 0 . 29 ) , which is a transmembrane receptor that controls a variety of developmental and physiological processes [32] . The polymorphisms in pastrel , CHKov1 and ref ( 2 ) P have highly specific effects , altering susceptibility to just one of the four viruses ( Figure 3 ) . Against these target viruses , the effect on the susceptibility of individual flies is considerable ( Figure 3 ) . Comparing flies that are homozygous for the resistant and susceptible alleles , the most significant SNP in pastrel increases survival times by 55% . The doc element insertion in CHKov1 reduces the proportion of infected flies by 39% , while the ref ( 2 ) P polymorphism is associated with a 24% reduction in infection rates ( see also [7] ) . When large numbers of statistical tests are performed and the statistical power is low , as is the case in many genetic association studies , there is a tendency to overestimate effect sizes [33] . However , the extremely low P-values associated with our resistance genes suggest our statistical power was high and therefore these effect size estimates are reliable . In pastrel there are six SNPs that are associated with resistance to DCV at P<10−12 . These include two adjacent SNPs in the 3′UTR ( genome positions: 3L:7350452 T/G , 3L:7350453 A/G ) , two non-synonymous SNPs ( 3L:7350895 Ala/Thr , 3L:7352880 Glu/Gly ) and two SNPs in introns ( 3L:7351494 C/T , 3L:7352966 T/G ) . All of these are in linkage disequilibrium , with the two SNPs in the 3′ UTR being perfectly associated ( these are therefore considered as a single variant in subsequent analyses ) . To try and disentangle which of the polymorphisms might be a causal variant , we fitted a general linear mixed-effects model in which all five variants were included as fixed effects . This allows us to calculate the marginal significance of each polymorphism ( i . e . the P-value after controlling for the effects of all the other SNPs ) . In this analysis only a non-synonymous SNP in the last coding exon of the pastrel remained highly significant ( 3L:7350895: F1 , 116 = 18 . 2 , P<0 . 0001; all other P-values>0 . 01 ) . This SNP occurred in 21 of 142 lines that were sequenced at this site . We then tested the significance of each of the other four variants individually while controlling for the effects of 3L:7350895 , by fitting general linear mixed-effects models and calculating sequential P-values from an ANOVA table . When we did this , all the other SNPs are significant ( P<0 . 0001 in all cases ) . If we assume that we have included all the polymorphisms in this region in our analysis , this suggests that the non-synonymous polymorphism 3L:7350895 and at least one of the other sites are causal variants — but strong linkage disequilibrium prevents us from identifying which one ( s ) . However , many polymorphisms , including indels , are missing from this dataset , so another polymorphism in this region that is not included in our analysis may be causing flies to be resistant to DCV . To confirm the antiviral role of pastrel , we used RNAi to knock down the gene in flies that were homozygous for the susceptible allele . To do this , we expressed hairpin RNAs that target the pastrel gene under the control of a constitutively and ubiquitously expressed Gal4 driver . When the flies were infected with a high dose of DCV , this resulted in a large reduction in survival rates relative to both a control with a similar genetic background ( Figure 4A; Cox proportional hazard mixed model: z = 3 . 62 , P = 0 . 0003 ) and a control where we knocked down a gene unrelated to viral resistance ( Figure 4A; Cox proportional hazard mixed model: z = 3 . 19 P = 0 . 001 ) . To allow us to investigate viral titres , we also infected flies with a lower dose of DCV , which caused less mortality ( Figure 4A ) . The viral titre in the flies where pastrel had been knocked down was ∼6 times greater than the background control and ∼15 times greater than the control gene ( Figure 4B; F2 , 18 = 23 . 3 , P = 10−5 ) .
We have found that a small number of major-effect polymorphisms can explain a substantial proportion of the genetic variation in the susceptibility of D . melanogaster to viral infection . These genes have either been previously identified by linkage mapping , or , in the case of pastrel , were verified by RNAi in this study . These polymorphisms are only seen when flies were infected with the two viruses that occur naturally in D . melanogaster populations — we were unable to detect any significant associations when using viruses that naturally infect other insects . The consequence of this is that the genetic variation in susceptibility to the naturally occurring viruses is substantially greater than to viruses from other species . Combined with previous data showing that two of these resistance alleles have been driven to a high frequency by positive selection [7] , [8] , [34] , these results suggest that selection by viruses in natural populations may be increasing genetic variation in disease susceptibility . As the resistance alleles that we detected have highly specific effects against a single virus , genetic variation in susceptibility to infection by viruses isolated from other species of insects has remained low . Our results support the suggestion that the genetic architecture of infectious disease susceptibility may be different from non-communicable diseases due to selection by parasites [13] , [35] . GWAS in humans have mostly focused on non-communicable disease , and have tended to find polymorphisms of modest effect . In contrast , work on infectious disease in humans has described numerous loci with a major-effect on susceptibility [3] , [13] , [35] , and similar patterns have been reported by QTL studies in other animals [36] . This has led to the suggestion that variation in pathogen resistance may often be controlled by a mixture of major-effect polymorphisms and other loci that are difficult to detect because they are rare or have small effects [13] . Our results corroborate this pattern , as while we find a few major-effect genes , over half of the total genetic variation remains unexplained . Furthermore , our results provide support for the role of natural selection by parasites in increasing the frequency and effect size of disease susceptibility loci . If this pattern proves to apply to other species , then GWAS on susceptibility to infectious disease promises to be a productive direction for future research . Parasites can result both in balancing selection maintaining polymorphisms in host resistance , and directional selection , which will ultimately fix the resistant allele [6] . Previous work has shown that the resistant alleles of CHKov1 and ref ( 2 ) P both arose recently by mutation and natural selection has caused them to increase in frequency [7] , [8] , [34] , [37] . Therefore , it appears as though directional selection is driving new resistance genes through the population , and this is increasing genetic variation in disease susceptibility . Directional selection on a trait can result in higher genetic variance when selection is acting on alleles that are initially at a low frequency [38]–[40] . If this is the case , selection will increase the frequency of rare alleles that previously contributed little to genetic variation in the population , and will therefore increase genetic variation in the trait [38]–[40] . For example , this process is thought to explain why selection by mate choice increases genetic variation in the cuticular hydrocarbons produced by Drosophila serrata [41] . In the case of the polymorphisms in ref ( 2 ) P and CHKov1 , previous work has suggested that they have undergone a ‘hard’ selective sweep , where selection has been acting on new or rare polymorphisms [7] , [8] , [34] , [37] . Therefore , they will have contributed little to genetic variation before selection , but now explain much of the heritability in this population . Certain traits , such as insecticide resistance , normally evolve in this way with selection acting on rare alleles [42] . This is thought to be because there are relatively few genetic changes that can cause insecticide resistance , and therefore there are too few mutations to generate much standing genetic variation [42] . If resistance to viruses also normally evolves due to selection on rare alleles , it may be common for directional selection to increase genetic variation . Fluctuating selection by parasites through time and space , or negatively frequency-dependent selection may also play an important role in increasing genetic variation . In addition to the genes that we have identified , the bacterial symbiont Wolbachia also makes D . melanogaster more resistant to viral infection [43] , [44] . Wolbachia occurs in natural populations , and protects flies against two of the viruses we studied — DCV and FHV [43] , [44] ( but not sigma viruses , Teixeira , Magwire and Wilfert , Pers . Comm . ) . As Wolbachia is transmitted vertically from mother to offspring , in many ways it can be regarded as another major-effect resistance polymorphism . We cured Wolbachia in our experiments with DCV and FHV and therefore have no data on its effects , but in natural populations of D . melanogaster it varies in prevalence from below 1% to near fixation [45] . This may affect how selection acts on the polymorphisms that we have identified , as our resistance genes may confer less of a benefit in populations where Wolbachia is common . Will the genetic architecture of resistance to other classes of pathogens be similar to the pattern we have seen for viruses ? In Drosophila , parasitoid wasps are one of the main causes of mortality in natural populations [46] , and both linkage mapping and artificial selection experiments suggest that resistance against these parasites is controlled by a few major-effect loci [18] , [47] , [48] . There is also extensive variation in bacterial resistance , and polymorphisms in immune system genes explain a substantial proportion of this variation [49] , [50] . It is difficult to compare these results directly to our own as comparatively few markers were genotyped and only known immune genes were investigated . Nonetheless , it would appear possible that bacterial resistance may have a more complex genetic architecture than virus resistance , involving more genes and epistatic interactions . Furthermore , there was not the clear difference between bacteria isolated from D . melanogaster and other organisms that we observed among our viruses [49] , [50] . This may be because there is a broad-spectrum induced immune response against bacteria [51] but not viruses [52] , or because the viruses may have a narrower host range , resulting in more rapid coevolution . In the long term , the spread of the resistance genes through populations could either result in the virus evolving to overcome host resistance , or a permanent increase in the levels of resistance seen in the host population . Due to their high mutation rates , short generation times and large population sizes , RNA viruses can evolve rapidly [53] . Therefore , it is perhaps unsurprising that during the 1980s and 1990s sigma virus genotypes that were not affected by ref ( 2 ) P resistance spread through European populations of D . melanogaster [54] . This suggests that we may be observing one side of a coevolutionary arms race between hosts and parasites . While the antimicrobial immune response of Drosophila is well-understood , we have only begun to understand in detail how Drosophila defends itself against viruses in the last six years [55] . Resistance to viruses could potentially evolve by altering the immune system ( antiviral genes ) , or host factors that are usually exploited by viruses during the viral replication cycle ( proviral genes ) . The only highly significant gene that we identified with a well-characterised function encodes ref ( 2 ) P , which is a homolog of the mammalian protein p62 [56] . This is a scaffold protein that has several functions , including in targeting cargoes such as protein aggregates and pathogens for destruction by autophagy — a process by which the cargo is wrapped in a double membrane vesicle called an autophagosome , which then fuses with the lysosome and is degraded [57] , [58] . Autophagy was recently found to be an important component of antiviral immunity in D . melanogaster infected with vesicular stomatitis virus ( VSV ) , which is another rhabdovirus that is related to DMelSV [24] , [59] . Therefore , it is possible that this polymorphism is affecting the antiviral immune response . We also found suggestive evidence that a polymorphism in scavenger receptor C1 may be important in defence against DAffSV . This gene functions both as a pattern recognition receptor of bacteria [30] and allows the uptake of dsRNA into cells , resulting in an RNAi response [31] . The function of this gene therefore suggests it is a strong candidate as a component of antiviral immunity . The functions of the two genes that have the largest effects on susceptibility , CHKov1 and pastrel , in antiviral defence remain unclear , although pastrel is thought to play a role in protein secretion [60] . Interestingly , knocking down the susceptible allele of this gene further increased susceptibility , suggesting that even the susceptible allele of the gene has some antiviral effect . Therefore , this gene may be part of the flies antiviral immune system . Characterising the role of CHKov1 and pastrel in the immune system or the viral life cycle promises to yield new insights into both how animals evolve resistance to infection , and how viruses interact with their hosts .
Many of the DGRP lines are infected with Wolbachia bacteria [28] , which affects susceptibility to DCV and FHV [44] . To clear the stocks of Wolbachia infection , flies were reared for two generations on food prepared by adding 6 ml of 0 . 05% w/v tetracycline to a vial containing 5 g instant Drosophila medium ( Carolina Biological , Burlington , North Carolina , U . S . A . ) and yeast . We checked the flies were uninfected by PCR as described in reference [61] . Lab fly stocks can also be naturally infected by DCV , so the lines were also cleared of natural virus infections by aging adult flies for 20 days and then dechorionating embryos with a ∼5% sodium hypochlorite solution [16] . A small number of the lines assayed for DAffSV were not treated in this way , but excluding these lines did not alter the results . Note that the DMelSV data was collected previously , before the lines were treated [7] ( Wolbachia does not affect sigma viruses , Teixeira , Magwire and Wilfert , Pers . Comm . ) . The generation prior to injecting the viruses , vials were set up containing two male and two female flies on cornmeal-agar food . For each fly line/virus combination , we set up 4 vials whenever possible . To maximise the level of cross-factoring , we always used different combinations of fly lines on each day of injecting . The flies were injected with 69 nl of the virus suspension intra-abdominally as described in detail in references [62] , [63] . The infection of D . melanogaster by DAffSV has not been characterised before , so we first injected flies with the virus and monitored viral titres and the characteristic symptom of sigma virus infection — sensitivity to CO2 . These pilot experiments confirmed previous results that the virus can replicate in D . melanogaster [62] , and also showed that infected flies are paralysed following exposure to CO2 ( Figure S3 ) . Therefore , to assay for infection by DAffSV , flies were exposed to pure carbon dioxide for 15 min at 12°C at 15 days post-injection . By 30 min post-exposure , flies are awake from this anaesthesia were classed as uninfected , but flies that were dead or paralysed were classed as infected [24] . To assay for susceptibility to DCV and FHV we recorded survival every 24 hours until all the flies had died . Due to a historical accident , the DCV and FHV experiments used male flies and the DAffSV and DMelSV experiments used females . Therefore , care should be taken in comparing these datasets ( note that these pairs of experiments also differ in the trait being measured ) . The DMelSV data has been published previously [7] , [8] . We have genotyped the DGRP lines for the polymorphisms in CHKov1 and ref ( 2 ) P by PCR as described in reference [7] , [8] . We knocked down the pastrel gene by RNAi . Males from line UAS-pst ( P{KK105159}VIE-260B ) were crossed to Actin-GAL4 ( w*;; P{GAL4-da . G32}UH1 ) virgins . As controls , males from a line with the same genetic background as UAS-pst ( y , w1118;P{attP , y+ , w3′} ) and UAS-CG10669 ( P{KK105150}VIE-260B ) were also crossed to Actin-GAL4 virgins . F1 females were injected with DCV . We used two doses of DCV ( high: TCID50 = 1000 and low: TCID50 = 690 ) . We note that these were from different viral preparations , which may explain why this small difference in TCID50 caused a large difference in mortality . We measured DCV titres by quantitative PCR using the SensiFAST™ SYBR & Fluorescein Kit ( Bioline , UK ) . DCV was amplified using the primers DCV 6060F ( 5′-CTTGCGGACCCTTTGTACGAC-3′ ) and DCV 7320R ( 5′-GCCATTCGAACTTGACCACGCAG-3′ ) . As an endogenous control we amplified Actin5c using the primers qActin5c_for2 ( 5′GAGCGCGGTTACTCTTTCAC 3′ ) and qActin5c_rev2 ( 5′ AAGCCTCCATTCCCAAGAAC 3′ ) . We performed three technical replicates of each PCR and used the mean of these in subsequent analyses . We calculated the titre of DCV relative to Actin5c as 2ΔCt , where ΔCt is the critical threshold cycle of DCV minus the critical threshold cycle of Actin5c . This approach assumes near 100% primer efficiency , which was confirmed using a dilution series of the template cDNA . We fitted a series of linear models to estimate genetic variances and covariances . For FHV and DCV , our data consisted of the lifespan of individual flies , which we treat as a Gaussian response in a general linear model . We fitted separate models for each virus , which were formulated as follows . Let yijk be survival time ( days after injection ) of fly k from line i and vial j . ( 1 ) where β is the mean survival time across all lines , bi is a random variable representing the deviation from the overall mean of the ith line , cj is a random variable representing the deviation jth vial from the line mean , and εi , j , k is the residual error . For DMelSV and DAffSV our data consist of numbers of infected and uninfected flies in each vial , which we treat as a binomial response in a generalized linear model . Let vi , k be the probability of flies in vial k from line i being infected . ( 2 ) where β is the overall mean , bi is a random variable representing the deviation from the overall mean of the ith line , and εik is a residual which captures over-dispersion within each vial due to unaccounted for heterogeneity between vials in the probability of infection . To estimate the genetic correlations between the viruses , we analysed data from all four viruses using a single model . To allow us to treat data from all four viruses as a binomial response in a generalized linear model , for FHV and DCV we used the numbers of dead and alive flies on a single day . Let vi , j , k be the probability of flies , in vial k from line i and infected with virus j being dead . ( 3 ) where β is a vector of the mean survival times of the four virus types , and xi is a row vector relating this fixed effect to vial i . bi , j is the random effect of virus k on line j , and was assumed to be multivariate normally distributed , allowing us to estimate separate line variances for each virus type and covariances between all pairwise combinations of viruses . εi , j , k is the residual which captures over-dispersion within each vial . The residuals were assumed to be normally distributed with a separate variance estimated for each virus type . The parameters of the models were estimated using the R library MCMCglmm [64] , which uses Bayesian Markov chain Monte Carlo ( MCMC ) techniques . Each model was run for 1 . 3 million iterations with a burn-in of 300 , 000 , a thinning interval of 100 and improper priors . We confirmed these results were not influenced by the choice of prior in the Bayesian analysis by also fitting models 1 and 2 using maximum likelihood ( data not shown ) . Credible intervals on variances , correlations , and heritability were calculated from highest posterior density intervals . As these fly lines are homozygous across most genes in the genome , the genetic variance , Vg , is half the between-line variance ( assuming additive genetic variation ) . This allows us to calculate the heritability of DCV and FHV as: ( 4 ) Where Vv is the between vial variance and Vr is the residual variance . As the DMelSV and DAffSV parameters are on a logit scale , we calculated heritability as: ( 5 ) where is the variance of a logistic distribution ( the cumulative distribution function of the logistic distribution is the inverse logit function , the link function used in the model; [65] , [66] ) . Note that the between-vial variance is included in Vr in this model . In Table 1 , we calculate Ve as Vr+ . When calculating the proportion of the heritability explained by the polymorphisms we identified , we recalculated Vg after accounting for these polymorphisms , and then adjusted the numerators of equations ( 4 ) and ( 5 ) accordingly . We also calculated the coefficient of genetic variation , CVg , for DCV and FHV as , where β is the mean survival time [27] . To estimate the proportion of the heritability that is explained by these genes we assumed the polymorphisms have additive effects , so their contribution to additive genetic variation is 2pqa2 , where p and q are the frequencies of the alleles in the population used to calculate h2 , and a is half the difference in the survival or infection probability of flies that are homozygous for the resistant and susceptible alleles ( i . e . a and −a are the genotypic values of the resistant and susceptible homozygotes ) . The maximum likelihood estimate of a , , was obtained by regressing genotype against the line means during the GWAS ( see below ) . An unbiased estimate of a2 was obtained as 2 minus the square of the standard error of . In this calculation , the heritability of resistance to DMelSV was recalculated from the line means which were treated as Gaussian data . We used robust statistics to analyse data on viral titres due to the presence of an outlier in the data . We fitted a linear model by robust regression using an M estimator and used a robust F test to assess significance [67] . To identify single nucleotide polymorphisms ( SNPs ) that were associated to susceptibility , we performed a GWAS using the published DGRP genome sequences [28] . We only included biallelic SNPs where the minor allele occurred in at least 4 lines , and treated segregating sites within lines as missing data . In the case of DCV and FHV , the susceptibility of the line was measured as the mean survival time of flies in each line ( with the vials of flies weighted by the number of flies in each vial ) . In the case of DMelSV and DAffSV , the susceptibility of the line was measured as the proportion of flies that were infected , as determined by the CO2 assay . The DAffSV data was arcsine square root transformed to remove the dependence of the variance on the mean . To each SNP we fitted the linear model ri , j = β+mi+εi , where ri , j is the susceptibility of flies with SNP genotype i from line j , β is the overall mean , mi is the SNP i genotype , and εik the residual . As major-effect polymorphisms affect the susceptibility of flies to DCV and DMelSV , the analysis was then repeated including the genotype of these genes as an additional explanatory variable . Because we are performing multiple correlated tests , we determined a genome-wide significance threshold for the association between a SNP and phenotype by permutation . The phenotype data were permuted over the different recombinant lines , the genome-wide association study was repeated as described above , and the minimum P-value across the entire genome was recorded . This was carried out 400 times to generate a null distribution .
|
In most animal populations , individuals vary genetically in how susceptible they are to infectious disease . To understand the genetic basis of this variation , we have infected a panel of inbred lines of the fruit fly D . melanogaster with viruses and have looked for genetic variants associated with resistance to infection . Using two viruses that naturally infect this species , we found a high level of genetic variation , much of which is due to a small number of genetic variants that have a large effect on virus resistance . Previous work has shown that two of these variants resulted from recent mutations that increased resistance and have been driven to a high frequency by natural selection . Furthermore , we did not find similar major-effect variants when we infected flies with viruses isolated from other species of insects . Therefore , selection for virus resistance appears to increase genetic variation in susceptibility to viral infection . Understanding the function of the genes , we have identified promises to give new insights into the antiviral defences of insects .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"genome-wide",
"association",
"studies",
"medicine",
"population",
"genetics",
"immunology",
"parasitology",
"population",
"biology",
"zoology",
"infectious",
"diseases",
"genetic",
"polymorphism",
"biology",
"genetics",
"of",
"the",
"immune",
"system",
"natural",
"selection",
"heredity",
"genetics",
"genomics",
"evolutionary",
"biology",
"genetics",
"of",
"disease",
"genetics",
"and",
"genomics"
] |
2012
|
Genome-Wide Association Studies Reveal a Simple Genetic Basis of Resistance to Naturally Coevolving Viruses in Drosophila melanogaster
|
Tsetse flies are important vectors of human and animal trypanosomiasis . Ability to reduce tsetse populations is an effective means of disease control . Lactation is an essential component of tsetse’s viviparous reproductive physiology and requires a dramatic increase in the expression and synthesis of milk proteins by the milk gland organ in order to nurture larval growth . In between each gonotrophic cycle , tsetse ceases milk production and milk gland tubules undergo a nearly two-fold reduction in width ( involution ) . In this study , we examined the role autophagy plays during tsetse fly milk gland involution and reproductive output . Autophagy genes show elevated expression in tissues associated with lactation , immediately before or within two hours post-parturition , and decline at 24-48h post-parturition . This expression pattern is inversely correlated with that of the milk gland proteins ( lactation-specific protein coding genes ) and the autophagy inhibitor fk506-bp1 . Increased expression of Drosophila inhibitor of apoptosis 1 , diap1 , was also observed in the milk gland during involution , when it likely prevents apoptosis of milk gland cells . RNAi-mediated knockdown of autophagy related gene 8a ( atg8a ) prevented rapid milk gland autophagy during involution , prolonging gestation , and reducing fecundity in the subsequent gonotrophic cycle . The resultant inhibition of autophagy reduced the recovery of stored lipids during the dry ( non-lactating ) periods by 15–20% . Ecdysone application , similar to levels that occur immediately before birth , induced autophagy , and increased milk gland involution even before abortion . This suggests that the ecdysteroid peak immediately preceding parturition likely triggers milk gland autophagy . Population modeling reveals that a delay in involution would yield a negative population growth rate . This study indicates that milk gland autophagy during involution is critical to restore nutrient reserves and allow efficient transition between pregnancy cycles . Targeting post-birth phases of reproduction could be utilized as a novel mechanism to suppress tsetse populations and reduce trypanosomiasis .
Tsetse are important vectors of disease caused by African trypanosomes , known as Sleeping Sickness in humans and Nagana in animals . An effective way to combat disease involves vector control applications , which are highly efficient due to the low reproductive output of tsetse . Tsetse flies are one of the few insects that employ viviparity [1 , 2] , characterized by the provision of nutrients beyond egg yolk to support embryonic development [3 , 4] and growth of larva to full term within the female uterus . Tsetse females produce a single mature third instar larva during each gonotrophic cycle following a 4–6 day period of intrauterine gestation [1 , 2] . Thus , these K-strategists produce only a modest 8–10 progeny per female , per lifetime . [1 , 5] . A critical adaptation underlying this reproductive strategy is the modification of the female accessory gland to secrete milk into the uterus for larval consumption . The milk is composed of proteins and lipids emulsified in an aqueous base [6–10] . During lactation , at least 6–10 mg of nutrients dissolved in 12–14 mg of water are transferred to the larva . Molecular characterization of tsetse milk revealed 12 major milk gland proteins , including Transferrin [10] , a Lipocalin ( Milk Gland Protein1 , MGP1 [9 , 10] ) , nine tsetse-specific milk proteins ( MGP2-3; [8–10] ) , and Acid Sphingomyelinase 1 ( aSMase1; [10] ) . In addition , peptides involved in mediating immune response , including Ubash3a and Peptidoglycan Recognition Protein-LB ( PGRP-LB , [10] ) , were identified as constituents of tsetse milk . Transcriptomic analysis revealed that the milk proteins represent over 47% of the total transcriptional output in lactating flies . The contribution of the milk protein transcripts declines to less than 2% of the total output two days after the lactation cycle , at parturition [10] . This drastic transcriptional change in milk protein expression indicates a rapid physiological change in the tsetse milk gland following birth . While previous studies have shown that milk gland involution after birth is quite rapid , with the milk gland shrinking to pre-lactation width in less than one day [11–14] , little is known about the mechanisms that underlie this process . Studies on milk gland ultrastructure have shown that lysosome density increases substantially immediately following birth , disappearing completely within two-to-three days [12 , 13] . Lysosomes are cellular indicators of milk gland autophagy , but little is known about their regulation and their role in the transition between lactating and non-lactating ( dry ) periods of the tsetse reproductive cycles . In this study , we examined the role of milk gland involution during the lactating to dry transition , specifically with respect to the autophagic mechanisms employed . We measured transcript abundance for multiple autophagy-related genes and performed knockdown studies on autophagy related gene 8a ( atg8a ) . Ecdysone application revealed the role of this hormone inducing post-parturition autophagy . Lastly , we used modeling to demonstrate how impaired autophagy could impact the rate of population growth . Our results suggest that autophagy in the milk gland is critical to facilitate the transition between the lactating and dry periods of the tsetse gonotrophic cycle , and that a delay in this autophagic regression reduces fecundity . Given that population reduction is an essential arm of vector control to decrease transmission of parasitic diseases to humans and animals , the ability to alter fecundity through interference with milk gland physiology can expand the tool box for tsetse fly suppression .
To assess the level of autophagy at different stages of the gonotrophic cycle , we examined images of the milk gland before lactation , during lactation and immediately after parturition , from previous publications [11–15] . During lactation , a three/five-fold increase in endoplasmic reticulum and Golgi apparatus is noted , while involution is characterized by a nearly 4-fold increase in the number of autolysosomes and lysosomes ( Fig 1A ) . The size of the milk gland reaches its maximum during the peak of milk production , when cells are filled with endoplasmic reticulum , Golgi apparatus , mitochondria , and the milk gland vacuole is fully replete ( Fig 1B ) . The milk gland quickly returns to its pre-lactation state within one day of parturition , following the proliferation of autolysosomes and lysosomes . To determine the role of autophagic mechanisms associated with milk gland involution , we measured transcript levels for genes whose products participate in or regulate autophagy [16 , 17] . The expression of autophagy related 1 , atg1 , ( involved in autophagosome induction [17] ) , atg6 ( involved in autophagosome nucleation ) [17] , and atg8a ( involved in autophagosome expansion ) [17] was measured during the period encompassing the peak of lactation through 48 hours following parturition ( Fig 2 ) . In addition , we monitored expression levels for fk506-bp1 ( an inhibitor of autophagy ) [18] , and Drosophila inhibitor of apoptosis 1 , diap1 ( a caspase inhibitor that suppresses apoptosis ) [19 , 20] . For comparison , we also show transcript levels for a major protein constituent of tsetse fly milk , milk gland protein 1 ( mgp1 ) , throughout the course of lactation and involution [10 , 21] . In whole female samples , autophagy genes ( atg1 , atg6 and atg8a ) show elevated expression during late lactation and parturition , immediately preceding the substantial reduction in mgp1 expression ( Fig 2 ) . Diap1 has a similar expression profile as atg genes , while fk506-bp1 expression is lowest at end of lactation and 24 hours following larviposition . To more precisely examine the expression of autophagy genes during lactation , we performed fine scale measurements of transcripts in the fat body/milk gland ( Fig 3 ) . Complete separation of the milk gland from the fat body into individual tissues is complicated by their intricate physical association [21] . Our results indicate that expression of mgp1 remains high immediately after parturition ( 0–2 hours ) , declines within 3–4 hours following parturition , and then begins to increase again at 120 hours post parturition upon initiation of the next cycle of larvigenesis ( Fig 3 ) . atg1 , atg6 , atg8a , and diap1 all follow a similar expression pattern in the fat body/milk gland . Expression of these genes remains low during the gonotrophic cycle , except at 0–2 to 7–8 hours after birth ( Fig 3 ) . The expression profile of the autophagy inhibitor fk506-bp1 is inverse of the autophagy genes; lowest expression was observed at 3–4 hours post-parturition ( Fig 3 ) . To validate expression in the milk gland , we performed in situ hybridization against atg8a , confirming atg8 expression in the milk gland following involution . No differences were noted in atg8a expression in the fat body between lactating flies and females immediately after birth ( 0-2h after parturition ) . This indicates that the observed differences in expression result from changes in the milk gland and not the fat body . We additionally examined ATG8 protein levels throughout the tsetse lactation cycle ( Fig 4 ) . Analysis of whole body total protein revealed that the ATG8a protein level is increased after birth ( post-parturition , Fig 4 ) . When preparations of fat body/milk gland were specifically examined , we observed a sharp increase in ATG8a protein 2-8h after birth , declining to undetectable levels by 20 hours post-parturition ( Fig 4 ) . These results correlate with the increase in transcript levels for autophagy-associated genes . To determine the role of ATG8a in milk gland involution , we employed siRNA injection to suppress atg8a expression ( Fig 5 ) . We were able to reduce atg8a expression by nearly 70% compared to control flies , injected with PBS or siGFP ( Fig 5 ) . This reduction was also noted at the protein level ( Fig 5 ) . Following atg8a knockdown , milk gland involution was substantially delayed . Knockdown flies required nearly three days for milk gland morphology to return to the pre-lactation state ( as determined by measurement of milk gland tubule diameter ) when autophagy was hindered via atg8a suppression . In control flies the entire process of involution is usually completed less than 24 hours after birth ( Fig 2B ) . Delayed involution is also evident according to mgp expression . mgp expression is enriched in siRNA injected flies one day after parturition , a time at when its expression would have otherwise declined to pre-lactation levels ( Fig 5 ) . In addition , there is a delay in the usual peak of mgp expression during the initiation of the subsequent lactation cycle ( Fig 5 ) . We examined fecundity following atg8a knockdown and observed a marked decrease in the number of progeny produced ( Fig 6A ) . Thus , fecundity reduction , measured as fewer progeny production per female , is partly a result of an extended pregnancy cycle due to the lack of optimal nutrients to support larval growth ( Fig 6B and 6C ) . Previous studies have documented a substantial increase in ecdysone titer immediately preceding parturition [22] , which declines to baseline levels within 24 hours . To determine if initiation of milk gland autophagy is associated with the high levels of ecdysone observed around parturition , we applied ecdysone at physiologically-relevant levels to whole females and observed a significant increase in both autophagy associated gene expression ( atg8a and atg6 ) and a reduction in mgp and asmase1 expression ( Fig 7A ) . This milk protein gene suppression correlates with a decrease in milk gland diameter following ecdysone treatment ( Fig 7B ) . Furthermore , ecdysone treatment increased the rate of larval abortion within 2d post treatment , and even more dramatically 3-4d after treatment ( Fig 7C ) . This delay likely results from impaired milk gland function , leading to larval malnutrition or starvation . These data suggest that the ecdysone peak immediately preceding larviposition likely contributes to the induction of milk gland autophagy that follows parturition . To assess the effect of involution associated autophagy disruption on tsetse population dynamics , we modeled the potential effects this would have on population growth following impaired autophagy following the first birth . Based on the modeling analysis , reproductive growth rate was reduced by 75% during the entire lifetime of a female when the process of milk gland involution was impaired following atg8a knockdown ( Fig 8A ) . The greatest reduction in reproductive growth rate occurred in early gonotrophic cycles ( cycles 2–5 ) , when female reproductive output is the highest ( Fig 8A ) . In addition , the population growth rates were reduced by ~15% following atg8a knockdown when compared to those injected with siGFP ( Fig 8B ) . Similar to the reduction in reproductive growth rate , the most substantial difference in the population growth rates occur during the early gonotrophic cycles . Overall , these modeling results indicate that impaired milk gland autophagy will reduce population maintenance of tsetse flies .
In this study we show that milk gland remodeling between cycles of lactation is critical to maintain optimum reproductive capacity in tsetse . Milk gland involution usually completes within one day after parturition . Expression patterns for autophagy genes suggest that milk gland remodeling likely relies on autophagy rather than apoptosis or necrosis . Suppression of autophagy and consequently involution delayed the subsequent pregnancy cycle by 2–3 days rather than 24 hours . An ecdysone peak that occurs immediately preceding parturition likely triggers autophagic processes in the milk gland . Hindering autophagy reduces the reproductive rate , which could lead to a decrease in population growth rate below replacement levels , indicating that this is a crucial aspect of the tsetse reproductive cycle . We have provided a summary illustrating the role of autophagy during involution during tsetse fly pregnancy ( Fig 9 ) . Autophagic mechanisms during insect reproduction have been examined in the mosquito , Aedes aegypti [23] . In many mosquito species , oogenesis is directly dependent on blood feeding [24 , 25] . Acquisition of a bloodmeal triggers vitellogenesis , during which the fat body produces a massive amount of yolk protein precursors that subsequently accumulate in the developing oocytes [24 , 25] . This process is regulated by ovarian ecdysiotropic hormone ( OEH ) , an insulin-like peptide produced in a subset of neurosecretory cells , which triggers the production of 20-hydroxyecdysone ( 20E ) [26–28] . An increase in the 20E titer , along with nutritional stimulation by amino acids through the target of rapamycin ( TOR ) signaling pathway , is responsible for the initiation and maintenance of vitellogenesis in mated , blood fed mosquitoes [29 , 30] . After oviposition , yolk protein gene expression is drastically reduced coinciding with an increase in lysosomal activity in the fat body [23] . Recently it was demonstrated that this rapid decline in expression of yolk proteins involves programmed autophagy that is regulated through ecdysone signaling [23] . In particular , a reduction in autophagy that occurs after vitellogenesis leads to a prolonged expression of Vg , a yolk protein gene , beyond what is necessary for egg production [23] leading to retardation of egg development in subsequent reproductive cycles [23] . This process is similar to that observed in tsetse , in which we have shown that autophagic regression of the milk gland is critical for subsequent reproductive cycles; impaired autophagy delays subsequent pregnancy cycles . While the effect reported here is moderate compared to the observations in A . aegypti [23] , it is clear that programmed autophagy , following the completion of the reproductive cycle , is a critical and conserved mechanism to ensure optimum fecundity for insects . Previous studies in tsetse revealed a rapid increase in circulating ecdysteroid immediately before parturition [22 , 31] , and application of ecdysteroid induces larval abortion in pregnant mothers [31 , 32] . Interestingly , it was an increase in ecdysone , rather than 20-hydroxyecdysone , that immediately preceded parturition [22 , 31] . This distinction is notable; in the uterus contractions are induced by ecdysone but not by 20-hydroxyecdysone [33] . Therefore , in this study we tested the effects of ecdysone on milk gland autophagy . Our results indicate that ecdysone induces the expression of autophagy genes in both whole body and decreases milk gland width . In mosquitoes , autophagy in the fat body at the termination of vitellogenesis requires ecdysteroid signaling , as suppression of the ecdysone receptor ( EcR ) was demonstrated to inhibit this process [23] . The observation of ecdysone mediated cyclical autophagic responses to changing reproductive states in both tsetse and mosquitoes suggests that this may be a conserved mechanism in Diptera . Recent studies have suggested that tsetse lactation is analogous to lactation in mammals as evidenced by a number of similarities between the two systems [10 , 34] . These similarities include clearly defined dry ( non-lactating ) and lactating periods [1] , transfer of beneficial bacteria via milk secretions [35 , 36] , aquaporin mediated water transfer for milk hydration within lactation associated tissues [7] and presence of functionally analogous proteins in tsetse and mammalian milk . These milk proteins include iron-binding proteins ( lactoferrin in mammals and transferrin in tsetse [37 , 38] ) , lipocalins and an expanded gene family coding for caseins in mammals and MGP2-10 in tsetse [10 , 39 , 40] that serve as both a protein resource and a critical source of phosphates . Following involution in mammals , tissue breakdown occurs via a combination of autophagic , apoptotic and necrotic mechanisms . The tissue is then repopulated with adipocytes until the next period of lactation [41 , 42] . Thus , the transition between dry and lactating periods in mammals involves a drastic shift in the type and amount of specific cells . For tsetse , milk gland involution likely does not involve apoptosis or necrosis . Rather , the process is limited to autophagic regression , likely because milk gland regeneration must begin within one day [1 , 12 , 14] , where a complete cellular breakdown and regrowth would require longer periods between bouts of lactation . This slower recovery after birth would lead to extended pregnancy , which in tsetse will reduce lifetime female fecundity , causing the population growth rate to drop below replacement levels [43 , 44] . Therefore , although tsetse and mammals share many analogous lactation-related phenomena , involution mechanisms differ due to the rapid rate at which tsetse must initiate subsequent lactation cycles to ensure fecundity . In conclusion , we provide the first analysis of the molecular mechanisms underlying milk gland involution . We show that milk gland involution is completed within the 24 hours following parturition and is subject to autophagic mechanisms . This likely occurs without cellular apoptosis or necrosis , allowing the cells to increase in size during milk production and undergo autophagic regression during involution . Interference with milk gland autophagy prevents a timely switch from lactating to nonlactating state , leading to lower reproductive output in subsequent generations , and a predicted failure to maintain population growth above what is necessary for population replacement . This project complements previous studies that showed that both dry [10 , 45] and lactating [45 , 46] periods are critical for efficient progeny generation , by defining the involution period as essential for the timely transition between lactation and dry states . Previous studies have suggested that interference with tsetse fly pregnancy represents a viable target to reduce tsetse populations [2 , 10 , 43] , through the targeting of milk production . This work further describes that interference with post-birth factors also represent targets for novel tsetse control strategies , which modeling suggests can suppress reproduction rates below what is required for population maintenance .
Glossina morsitans morsitans were reared at Yale University and supplemented with those from the Slovak Academy of Sciences . Flies were maintained on bovine blood meals provided through an artificial feeding system at 48h intervals . Tissue samples were collected from pregnant females ( 16-18d after adult emergence ) carrying third instar larvae and at multiple intervals following parturition . Images of milk gland cells were acquired with permission from several studies describing the changes in these cells throughout lactation and birth [11–14] . Each image was divided into ten equal quadrants and the area occupied within the milk gland cells by endoplasmic reticulum , Golgi apparti , and autolysosome/lysosomes was measured . Milk gland diameter was measured throughout the course of pregnancy by removing the fat body , uterus and milk gland at various points during the pregnancy cycle . These organs were maintained in PBS ( pH 7 . 4 ) for less than one hour on ice , after which the diameter of the milk gland was assessed microscopically . RNA and protein isolations were performed using Trizol reagent on whole flies and milk gland/fat body samples following instructions provided by the manufacturer ( Invitrogen ) . RNA was cleaned with an RNeasy Mini Kit ( Qiagen ) . Complementary DNA was synthesized using a Superscript III reverse transcriptase kit from 1μg of the total RNA isolated from each sample . Transcript levels for atg8a , mgp , atg1 , atg6 , fk506-bp1 , and diap1 ( gene sequences acquired from Glossina genome project , vectrobase . org ) were determined via qPCR by employing the CFX real-time PCR detection system ( Bio-Rad , Hercules , CA ) with primers specific to each target gene ( S1 Table ) . All readings were obtained on four biological replicates that were normalized to tsetse tubulin expression levels . CFX Manager software version 3 . 1 ( Bio-Rad ) was used to quantify transcript expression of each gene and conducted according to methods developed in previous studies [7 , 34] . Short interfering RNA ( siRNA ) comprised of two duplex sequences ( UAAUACGACUCACUAUAGGGACAACGUCAUUCCACCAACA and UAAUACGACUCACUAUAGGGGCCCAGAAAGGGUGUGAAUA ) targeting atg8a were purchased from IDT in Coralville , IA . Green fluorescent protein ( GFP ) targeted siRNA ( GAUGCCAUUCUUUGGUUUGUCUCCCAU and CUUGACUUCAGCACGUGUCUUGUAGUU ) was used as a control . Previous studies have shown that gene knockdown using siRNA injection yields robust suppression . A spectrophotometer was used to ensure the concentration of each siRNA ( GFP and atg8a ) was adjusted to 700–750 ng/μl . siRNA ( ~1 . 5μl ) was injected into the thorax of tsetse mothers harboring a second instar larva in the uterus . Importantly , previous studies have shown that siRNA injection into the mother does not affect larval transcript levels [7 , 10] and as such , any deviations seen in the larva are due to maternal knockdown , not to the unintentional suppression of larval genes . All expression levels were normalized to tubulin . Transcript levels were assessed via qPCR as previously described . Proteins for each time point were extracted in three groups of five flies according to previously described methods [15 , 34] . Anti-sera used Include: Rabbit αTubulin ( GmmTub , 1:1000 ) and rabbit α ATG8a ( 1:1000 ) were from Bryant and Raikhel [23] . 1/400 of a fly was loaded into each well . Blots were blocked overnight in blocking buffer ( PBS , 3% BSA and 0 . 5% Tween 20 , pH 7 . 4 ) . Exposure time to anti-sera was taken from Attardo et al . [21 , 36] . Supersignal West Pico Substrate ( Pierce , Wobrun , MA ) was used to visualize each blot on an Image Station 2000R ( Kodak , New Haven , CT ) . Milk gland tubules intertwined with fat body were collected from mothers 0–2 hours after parturition . The combined milk gland/fat body was placed into Carnoy’s fixative for a five-six day fixation period [36] . Antisense/sense digoxigenin-labeled RNA probes were generated using the MAXIscript T7 transcription kit following manufacturer's protocol ( Ambion , Austin , TX ) using a primer set with a T7 primer ( S1 Table ) [36] . Antibody solutions were made using α -Digoxigenin-rhodamine Fab fragments ( Roche ) for FISH probe detection ( 1:200 dilution ) and rabbit α -GmmMGP ( 1:2500 ) antibodies [21 , 36] . Alexa Fluor 488 goat α -rabbit IgG ( Invitrogen ) at a dilution of 1:500 was added as a secondary antibody for immunohistochemistry [36] . Slides were mounted in VECTASHIELD Mounting Medium with DAPI ( Vector laboratories Inc . Burlingame , CA ) . Samples were observed using a Zeiss Axioskop2 microscope ( Zeiss , Thornwood , NY ) equipped with a fluorescent filter and viewed and imaged at 400x magnification . Images were captured using an Infinity1 USB 2 . 0 camera and software ( Lumenera Corporation , Ottawa , Ontario , Canada ) and merged in Adobe Photoshop . Ecdysone injections were performed according to methods previously described for tsetse flies [32 , 47] , with modification . Flies used for injection harbored a late first instar or second instar larva , and were in their second reproductive cycle ( 4–5 days before birth ) . This is a period when the expression of autophagy genes should be extremely low ( This study ) . Edysone ( Sigma-Aldrich ) was diluted into 95% ethanol , creating a stock solution of 10 μg/μl . The stock solution was diluted to 1 ng/μl with PBS on the day of injection . Each fly was injected with 0 . 5 μl . This injection amount is physiologically-relevant based on the increase in ecdysone to ~150 pg/μl noted immediately before birth [22] . RNA was extracted from female flies 24 hours after ecdysone injection , as before . Milk gland width was examined 24 hours after ecdysone treatment . The number of abortions was monitored for three days following ecdysone injection . Estimation of the impact of impaired autophagy on population growth , was determined by utilizing a simple model for tsetse population growth modified from Michalkova et al . [43] . The model population for impaired involution-associated autophagy was parametrized with data from the siATG8a treatments . The non-impaired ( control ) model population was parametrized using data from the siGFP treatment ( S2 Table ) . In each model , mortality was assumed to be normal levels associated tsetse population [48 , 49] . We calculated the mean and standard deviation and gonotrophic-cycle length for the control and autophagy-impaired groups over the course of 12 gonotrophic cycles ( S2 Table ) . We modeled fecundity Fjk as a beta random variable with parameters chosen to match the mean and standard deviation of the data . Similarly , gonotrophic-cycle length tjk was modeled as a log normal random variable with parameters chosen to match the data for each gonotrophic cycle . Given values of fecundity and gonotrophic cycle length , number of female offspring produced by a single female tsetse over its lifetime is Rj = p∑Fjk Sjk , where p is the probability that a deposited pupa is female , which we took to be 55% [48] and Sjk is the survival , the probability of surviving to gonotrophic cycle k . We modeled survival as Sjk = SpupasTjk , where Spupa is the probability that a deposited pupa survives to emerge as an adult , which we took to be a conservative number at 85% [48 , 49]; s is the probability of surviving each adult day , which we took to be 98% [48 , 49] , and Tjk = ∑tjk is the number of days from emergence until the end of gonotrophic cycle k . The population growth rate is then rj = RjD , with the generation time defined to be D = Dpupa + Dadult; Dpupa is the mean duration of the pupal stage , which we took to be 31 . 4 days [49]; and Dadult = −[log ( s ) ]−1 is the mean adult lifespan . We calculated the population growth rate rj for each treatment group and the difference in growth rate between the two treatment groups , r1 – r2 , for 10 , 000 samples of our model .
|
Tsetse flies are vectors for trypanosomes that cause both African sleeping sickness in humans and Nagana in animals . The reduction of tsetse populations is the most efficient way to reduce the prevalence of this economically important disease with current control methods including pesticide application , traps , and sterile insect techniques . Tsetse pregnancy and milk production represent a species-specific target for population control and milk gland transition during each larval growth cycle could represent a novel target for tsetse control . Within one day after birth , the milk gland organ , essential for provisioning nutrients to the intrauterine larva , undergoes involution marked by an ecdysone driven increase in autophagy that allows breakdown of this gland . Inhibiting the process of autophagy prevents the timely transition from the lactation phase to the dry phase , triggering a delay in subsequent pregnancy cycle . This misregulation of milk gland involution leads to an overall decrease in the number of offspring that each female can produce per lifetime . This study has determined the molecular components of this process , and reveals new targets of interference for vector control .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"cell",
"death",
"medicine",
"and",
"health",
"sciences",
"milk",
"autophagic",
"cell",
"death",
"body",
"fluids",
"maternal",
"health",
"obstetrics",
"and",
"gynecology",
"gene",
"regulation",
"cell",
"processes",
"diet",
"reproductive",
"physiology",
"endocrine",
"physiology",
"nutrition",
"women's",
"health",
"pregnancy",
"population",
"biology",
"small",
"interfering",
"rnas",
"lipids",
"fats",
"endocrinology",
"gene",
"expression",
"lactation",
"population",
"metrics",
"biochemistry",
"rna",
"anatomy",
"cell",
"biology",
"nucleic",
"acids",
"beverages",
"physiology",
"genetics",
"biology",
"and",
"life",
"sciences",
"non-coding",
"rna",
"population",
"growth"
] |
2018
|
Rapid autophagic regression of the milk gland during involution is critical for maximizing tsetse viviparous reproductive output
|
When the primary visual cortex ( V1 ) is damaged , the principal visual pathway is lost , causing a loss of vision in the opposite visual field . While conscious vision is impaired , patients can still respond to certain images; this is known as ‘blindsight’ . Recently , a direct anatomical connection between the lateral geniculate nucleus ( LGN ) and human motion area hMT+ has been implicated in blindsight . However , a functional connection between these structures has not been demonstrated . We quantified functional MRI responses to motion in 14 patients with unilateral V1 damage ( with and without blindsight ) . Patients with blindsight showed significant activity and a preserved sensitivity to speed in motion area hMT+ , which was absent in patients without blindsight . We then compared functional connectivity between motion area hMT+ and a number of structures implicated in blindsight , including the ventral pulvinar . Only patients with blindsight showed an intact functional connection with the LGN but not the other structures , supporting a specific functional role for the LGN in blindsight .
Damage to the primary visual cortex ( V1 ) that may occur following a stroke causes visual loss in the corresponding part of the visual field ( homonymous hemianopia , [1] ) . However , extensive research has shown that some patients retain an ability to respond to images inside their scotoma , even though they may not consciously see them [2] . This phenomenon is called blindsight , and recent work applied diffusion MRI and tractography in patients with V1 damage to try to uncover which pathways may underlie this residual visual function [3] . A connection between the lateral geniculate nucleus ( LGN ) and human motion area , hMT+ , was found to be intact in patients with blindsight but was absent or impaired in patients without blindsight . The other pathways tested , which included a connection between hMT+ and the superior colliculus ( SC ) , and with hMT+ in the opposite hemisphere , did not show this pattern . Unfortunately , a limitation of diffusion MRI is that it investigates purely structural connections , which may not relate directly to the function under investigation [4] . Furthermore , seed-based tractography is restricted to pathways chosen by investigators ( see also [5–7] ) , which in this case did not include a connection with the thalamic pulvinar nucleus . Neither a specific role for the LGN nor a functional connection to hMT+ has been shown in human blindsight and would significantly advance our understanding of how patients respond to visual images in the absence of V1 . The current study investigated behavioural and functional MRI responses to speed of motion in a group of patients with V1 damage in adulthood ( n = 14 ) , and healthy age-matched controls ( n = 8 ) . Patients were categorised as blindsight positive or negative according to their ability to detect the visual stimulus within their blind visual field . We then compared measures of activity and functional connectivity between the two patient groups and healthy controls . Patients with blindsight demonstrated significant fMRI activity in hMT+ in the damaged hemisphere , with a relatively preserved hMT+ response to speed in the blind hemifield . Critically , patients with blindsight also showed intact functional connectivity between hMT+ and LGN in the damaged hemisphere , which was absent in patients without blindsight . This was specific to the LGN , as both patient groups demonstrated preserved functional connectivity between hMT+ and ( i ) ventral pulvinar , ( ii ) SC , and ( iii ) contralateral hMT+ , which was no different than in healthy controls . The pulvinar , in particular , is implicated in human and nonhuman primate studies in which V1 is damaged early in life [8 , 9]; however , this region is yet to be evaluated in adult-onset blindsight . Our findings support a critical functional role for the LGN and its specific connection with hMT+ in adult human blindsight , reinforced by recent evidence for an intact anatomical connection between these structures [3] .
Blindsight-positive patients demonstrated significant fMRI activity in contralateral hMT+ for moving versus static dots in the blind hemifield ( Fig 1Ai and 1Aii ) . Overall hMT+ blood oxygen level–dependent ( BOLD ) signal change for all 5 speeds was significantly greater than baseline ( Fig 1Aiii , t = 3 . 0 , p = 0 . 02 , df = 7 ) . Activity was somewhat lower in intensity and spatial extent than the sighted field or healthy controls ( Fig 1C ) , but there was no significant difference between blind and sighted responses in blindsight-positive patients ( Fig 1Aiii t = 1 . 3 , p = 0 . 2 , df = 39 ) . Whilst the absence of a significant difference may result from a lack of power ( n = 8 ) , blindsight-negative patients with a lower ‘n’ ( Fig 1B ) showed a marked effect of hemifield ( paired t = 2 . 9 , p = 0 . 007 , df = 29 ) and no demonstrable hMT+ activity for moving versus static dots in the blind hemifield ( Fig 1Bi and 1Bii ) nor averaged across all conditions ( Fig 1Biii , t = 0 . 6 , p = 0 . 6 , df = 5 ) . In addition to hMT+ activity , blindsight-positive patients with right V1 damage ( Fig 1Aii , n = 2 ) showed activity in Jülich-defined right V2 and left V4 , although there was no such activity in patients with left V1 damage ( n = 6 ) . In blindsight-negative patients , visual-evoked responses also appeared within small regions of ipsilesional V4 and V2 , as well as the occipital pole and inferior parietal lobule , implying that activity in these regions was insufficient for motion perception . It was noteworthy that overall hMT+ signal change in the undamaged hemisphere was also slightly reduced in both patient groups compared to age-matched controls . A possible explanation is that unilateral V1 damage can negatively impact sighted processing in the opposite hemisphere , perhaps via a disturbance of interhemispheric interactions [10 , 11] . Speed of motion ( 0°–32°/s ) had a significant impact on hMT+ activity in the sighted hemifield of patients and equivalent left hemifield of controls ( 2-way ANOVA: F ( 4 , 14 ) = 5 . 1 , p < 0 . 001 ) , with no effect of participant group ( Fig 1iv , F ( 2 , 14 ) = 2 . 2 , p = 0 . 12 ) . In the blind hemifield of patients ( right hemifield of controls ) there was a similar effect of speed ( F ( 4 , 14 ) = 3 . 5 , p = 0 . 01 ) but also group ( F ( 2 , 14 ) = 7 . 4 , p = 0 . 001 ) . The pattern of responses in blindsight-positive and -negative patients differed markedly . Blindsight-positive patients showed a positive relationship between hMT+ signal change and speed ( r = 0 . 8 , 5 speeds ) , whilst the correlation coefficient was negative in blindsight-negative patients ( r = −0 . 6 ) . This difference was not simply driven by a difference between motion and static responses , as excluding the static conditions increased the significance even further ( r = 0 . 67 versus r = −0 . 99 , z = −2 . 44 , p = 0 . 01 ) . Specifically , blindsight-positive patients showed a relatively ‘normal’ hMT+ relationship with speed in the blind hemifield that was similar to the pattern in healthy controls ( Pearson r = 0 . 89 , p = 0 . 04 , 5 speeds ) and patients’ own sighted hemifield ( r = 0 . 98 , p < 0 . 01 ) . This was not the case for blindsight-negative patients , either when compared to controls ( r = 0 . 21 , p = 0 . 7 ) or to their own sighted hemifield ( r = 0 . 29 , p = 0 . 6 ) . To determine how activity in hMT+ correlated with subcortical activity , we examined the fMRI time series after stimulus-evoked responses had been regressed out . Specifically , we compared the residual pattern of activity in hMT+ with ( i ) LGN , ( ii ) ventral pulvinar , ( iii ) SC , and ( iv ) hMT+ in the opposite ( undamaged ) hemisphere , using subject-specific regions of interest ( ROIs ) ( S3 Fig ) . We also performed whole-brain analyses to measure the voxels where neural activity most closely matched the time series of hMT+ and these subcortical structures . LGN–V1 correlation in the nondamaged hemisphere ( Fig 2A ) was similar across all participant groups ( F = 1 . 0 , p = 0 . 4 , df = 2 ) . Bilateral hMT+ correlation was also very similar across all groups , indicative of preserved functional connection between hMT+ in patients irrespective of blindsight status ( Fig 2B , F = 1 . 9 , p = 0 . 2 , df = 2 ) . There was a significant effect of participant group on LGN–hMT+ correlation ( Fig 3A , F = 10 . 1 , p = 0 . 001 , df = 2 ) . Pairwise analysis for the damaged hemisphere showed that blindsight-negative patients had a significantly lower mean correlation coefficient compared to blindsight-positive patients ( −0 . 03 ± 0 . 08 SE versus 0 . 25 ±0 . 03 SE , t = 3 . 6 , p = 0 . 003 , df = 12 ) and remained at zero . In contrast , functional connectivity between hMT+ and ventral pulvinar ( Fig 3B ) or hMT+ and SC ( Fig 3C ) showed no effect of group ( F = 0 . 5 , p = 0 . 6 , df = 2 , for both ROIs ) and no difference between blindsight-positive and -negative patients in the damaged hemisphere ( pulvinar: t = 0 . 9 , p = 0 . 4 , SC: t = 0 . 5 , p = 0 . 6 , df = 12 ) . Blindsight-negative patients also showed a hemispheric difference for LGN ( t = 3 . 0 , p = 0 . 01 , df = 10 ) but not ventral pulvinar connectivity ( t = 2 . 1 , p = 0 . 06 , df = 10 ) . This suggests that the key difference in functional connectivity between blindsight-positive and -negative patients was the presence of a functional connection between hMT+ and LGN in the damaged hemisphere . We performed an additional analysis without regressing out stimulus-evoked responses and found the same results ( one-way ANOVA F = 4 . 3 , p = 0 . 02 , df = 2; paired t = 3 . 4 , p = 0 . 005 , df = 12 ) , implying that in patients with blindsight this functional connection was also present during visual function . These analyses , however , have not evaluated whether differences in connectivity were specific to our predefined regions of interest or if they reflect a global process independent of the ROIs and our hypothesis . To address this , we performed an additional whole-brain mixed effects analysis measuring the voxels where neural activity most closely matched the time series of LGN , ventral pulvinar , and hMT+ . This technique was used to generate seed region correlation maps , in which the ‘seed’ ROI should necessarily demonstrate a high correlation coefficient represented by a beta of 1 [12] . Co-active regions would similarly possess a high beta , with maps retaining a high spatial resolution since every voxel is tested [13] . As expected , group correlation maps showed a high beta in the ‘seed’ regions , reflecting consistency between subject-specific ROIs and their precise transformation to standard space ( Fig 4 ) . When LGN was the ‘seed’ region , control participants also demonstrated a relatively high beta in a small region of the calcarine cortex corresponding to retinotopically active V1 ( Fig 4Aiv and 4Civ ) , reflecting the functional geniculostriate pathway . A similar region of calcarine cortex was co-activated in the undamaged hemisphere of patients when the seed was LGN in the same hemisphere ( Fig 4Cv and 4Cvi ) . For LGN in the damaged hemisphere of both patient groups , there was no demonstrable V1 co-activation ( Fig 4Av and 4Avi ) , likely reflecting the damage to that region ( and/or its input ) . In contrast , blindsight-positive patients showed a relatively high beta in hMT+ of the damaged hemisphere compared to other participant groups and a small region of calcarine cortex in the undamaged side ( Fig 4Aii ) . In contrast to the LGN , when ventral pulvinar was used as the ‘seed’ , all groups showed robust co-activity in the SC , an area known to share an important connection with the pulvinar [14 , 15] . There was also notable connectivity with V1 in the undamaged hemisphere ( Fig 4D , [16] ) but no major connectivity with hMT+ ( Fig 4B ) . When hMT+ was used as the ‘seed region’ ( Fig 5 ) , controls demonstrated marked functional connectivity throughout the visual cortex , including hMT+ in the opposite hemisphere and V1 in both hemispheres , consistent with previous reports [17] . In subcortical regions , connectivity was also demonstrable in the LGN bilaterally , albeit to a lesser extent ( Fig 5A and 5B , left column ) . No equivalent connectivity was seen in the ventral pulvinar or SC , although these regions were co-activated in both hemispheres of all three participant groups if a slightly lower beta threshold of 0 . 32 was used ( rather than 0 . 35 ) . This pattern of connectivity was likely to reflect the major visual pathway and its rich network of intra- and interhemispheric connections . A very similar pattern was demonstrated in blindsight-positive patients , except for a relatively low beta in calcarine cortex of the affected hemisphere , reflecting the region of tissue damage . This was less apparent in blindsight-negative patients . Blindsight-negative patients also showed relatively poor connectivity with LGN , particularly in the damaged hemisphere , where ipsilateral geniculate co-activity was not demonstrable even when a low threshold was applied ( beta > 0 . 2 ) . Of the eight blindsight-positive patients , only half could discriminate direction of motion above chance . We were interested in whether this subgroup showed greater BOLD activity and/or connectivity compared to blindsight-positive patients who were unable to discriminate motion direction . Indeed , these four patients did show slightly stronger motion responses in hMT+ ( 0 . 3 ± SE 0 . 13 versus 0 . 1 ± 0 . 07 , t = 1 . 6 , p = 0 . 2 , df = 6 ) , as well as slightly greater LGN/hMT+ functional connectivity ( 0 . 27 ± 0 . 05 versus 0 . 22 ± SE 0 . 06 , t = 0 . 6 , p = 0 . 5 , df = 6 ) , although the differences were relatively small and nonsignificant . To investigate the behavioural–neuroimaging association further , we performed a correlation analysis between mean behavioural performance in both experiments and fMRI activity and connectivity across all patients ( n = 14 ) . hMT+ motion responses showed a weak but positive correlation with behavioural performance ( r = 0 . 38 , p = 0 . 18 ) , as did LGN–hMT+ connectivity ( r = 0 . 38 , p = 0 . 18 ) . There was also a positive but weaker correlation between behavioural performance and pulvinar–hMT+ ( r = 0 . 29 , p = 0 . 31 ) or SC–hMT+ connectivity ( r = 0 . 21 , p = 0 . 47 ) . More notable was a significant correlation between behavioural performance and the ratio of blind to sighted fMRI responses in hMT+ ( r = 0 . 62 , p = 0 . 018 ) . In other words , patients with the strongest hMT+ activity for motion in the blind field relative to their sighted field performed best at behavioural assessments of blindsight using the same visual stimuli . In order to determine the extent to which blindsight performance and the underlying neural mechanisms relate to the size of the lesions , we quantified the damage in each patient . Lesion size did not differ significantly between blindsight-positive and -negative groups ( t = 1 . 5; df = 12; p = 0 . 16 ) , although blindsight patients on average had smaller lesions ( 13 , 693 mm3 ± 1 , 825 mm3 SEM ) than those without blindsight ( 21 , 212 mm3 ± 5355 mm3 SEM ) . This is illustrated by the summed lesion masks for each patient group on a standard space template ( S2D and S2E Fig ) and the individual lesion maps in structural space ( S4 Fig ) . Reflecting the small difference in lesion size between the two patient groups , there was a moderate inverse relationship between lesion size and behavioural performance across both tasks ( r = −0 . 48; p = 0 . 08 ) . However , there was no relationship between lesion size and functional LGN–hMT+ connectivity ( r = 0 . 21 , p = 0 . 5 ) , or hMT+ signal change ( r = −0 . 24 , p = 0 . 4 ) . Thus , lesion size did not appear to be a critical factor in determining neural response . Aside from lesion size , the degree to which lesions involve hMT+ and its innervating connections would also be critically important . Lesion maps ( S4 Fig ) suggest that the lesion of one blindsight-negative patient ( P12 ) did encroach upon hMT+ , while it appeared intact in the other five blindsight-negative patients , including its afferent white matter . P1 also had a large lesion extending to the posteromedial border of hMT+; however , this should not have impacted upon the subcortical or interhemispheric white matter connections ( see S4 Fig for map ) . Accordingly , we quantified lesion involvement of hMT+ and its surrounding white matter in each patient ( S5 Fig ) . We found no difference when comparing blindsight-positive and -negative patients ( t = 1 . 7 , p = 0 . 11 , df = 12 ) , and hMT+ lesion size showed no association with LGN–hMT+ functional connectivity ( t = 1 . 6 , p = 0 . 13 , df = 13 ) , hMT+ activity ( t = 1 . 4 , p = 0 . 19 , df = 13 ) , or behavioural performance ( t = 1 . 6 , p = 0 . 13 , df = 13 ) in paired t tests using lesion size as the dependent variable . When recruiting a relatively large group of patients , it is challenging to ensure a completely homogenous lesion pattern . However , any specific differences in lesion size or extent can be informative . P12 , whose lesion may have encroached on hMT+ , demonstrated weak functional connectivity ( outside one SD of the mean ) in all three subcortical pathways ( pulvinar–hMT+: r = 0 . 004 , SC–hMT+: r = −0 . 13 ) and between hMT+ bilaterally ( r = 0 . 33 ) , and this contributed to slightly lower averages for pulvinar , collicular , and interhemispheric ( but not LGN ) connectivity in blindsight-negative patients ( see Fig 2B and Fig 3B and 3C ) . The variability amongst naturally occurring human V1 lesions has been highlighted as a limitation of human research compared to nonhuman primates [18] . It certainly emphasizes the limitations in carrying out individual case studies , which have predominated in the blindsight literature over the last several decades . However , the heterogeneity in the precise location of structural damage can also be extremely useful and has permitted patients to be classified according to their distinct residual visual performance . By determining which connections and characteristic fMRI responses are consistent amongst blindsight-positive and -negative patients , it may be possible to identify which underlying structures and pathways are involved . To summarise our results , blindsight-positive patients showed ( i ) significant neural activity in hMT+ to motion stimuli in the ‘blind’ visual field , ( ii ) a relatively preserved response to speed in hMT+ , and ( iii ) a correlation between resting BOLD signal in LGN and hMT+ in the damaged hemisphere . None of these findings were demonstrable in patients without blindsight . However , blindsight-negative patients did show functional connectivity between hMT+ and ( i ) SC and ( ii ) ventral pulvinar that was no different to healthy controls or patients with blindsight , suggesting that the LGN has a specific functional role in blindsight .
In summary , we identified a functional connection between LGN and hMT+ in patients with blindsight that was absent in patients without blindsight , despite a retained functional connection with ventral pulvinar and SC . This supports a critical functional role for the LGN in human blindsight , and in particular its connection with hMT+ , reinforced by recent evidence for an intact anatomical connection between these structures [3] . Our results also revealed that hMT+ does not require intact V1 for a normal speed response , although it does require a functional connection with the LGN . This suggests that the LGN may support motion-selective input to hMT+ in the absence of V1 . These results focus on behavioural and neural responses to visual motion , which is a critical component of blindsight ( see [66] for recent review ) . In future work , it will be necessary to explore how such pathways interact with other aspects of blindsight function and whether distinct tasks or stimuli might engage separate mechanisms in the absence of V1 .
Fourteen patients with adult-onset unilateral V1 damage took part in this study ( see S1 Table for details ) . The location of any additional non-V1 damage is shown in S2D and S2E Fig , and S4 Fig . No patients sustained damage to subcortical structures , including the LGN and pulvinar . Average age at the time of participation was 55 . 6 years ± 15 . 2 SD; average time after pathology onset was 49 months ( 6–252 months ) . Eight age-matched , healthy participants ( 50 . 1 ± 14 . 6 SD years ) served as controls . Written informed consent was obtained from all participants , and ethical approval was provided by the Oxford Research Ethics Committee ( Ref B08/H0605/156 ) . All experiments adhered to the Declaration of Helsinki . Visual stimuli were generated using MATLAB ( Mathworks ) and the Psychophysics Toolbox [67 , 68] . Each dot was 0 . 075° in diameter and had an infinite lifetime , with an average dot density of 8 dots/°2 . Visual stimuli consisted of an aperture of 5° or 8° diameter containing static or coherently moving black dots ( luminance 0 . 5 cd/m-2 ) at 4 , 8 , 20 , or 32°/s on a uniform grey background of luminance 50 cd/m-2 . Stimuli were positioned inside a region of dense visual field loss in patients a minimum of 2 . 5° from fixation ( S6 Fig ) . The extent to which stimuli covered the scotoma ( as a percentage ) was estimated for each patient from the Perimetry Visual Field Index ( VFI ) , provided in S6 Fig . Stimulus size and position was matched as closely as possible in eight age-matched controls ( S2A Fig ) , with no significant difference in distance between fixation and stimulus edge ( x or y coordinates ) when comparing patients to controls ( mean x = 3 . 6 ± 0 . 30 SE patients versus 3 . 8 ± 0 . 4 SE controls , t = 0 . 4 , p = 0 . 7 , df = 20; mean y = 0 . 65 ± 0 . 66 SE patients versus 0 . 49 ± 0 . 68 SE controls , t = 0 . 2 , p = 0 . 9 , df = 20 ) . Stimuli in blindsight-negative patients were also no deeper into the visual field than blindsight-positive patients ( stimulus edge 4 . 0° ± SD 0 . 9 in blindsight-positive versus 3 . 5° ± SD 1 . 2 in blindsight-negative patients , t = 0 . 9 , p = 0 . 4 , df = 12 ) . To select the stimulus location in patients , we required a perimetry threshold p < 0 . 005 or < −20dB ( which ever was more stringent ) for pattern deviation compared to age-matched controls at the stimulus location . This meant that the patients in our study were unable to see even the brightest unattenuated stimuli at that location in the visual field . To verify that we had not inadvertently chosen locations in blindsight-positive patients that were more sensitive than those in blindsight-negative patients , we calculated the average pattern deviation by taking the value closest to the stimulated location using Humphrey Perimetry . The residual visual sensitivity was no different in the two groups ( −32 . 8dB ± SE 0 . 8 blindsight positive versus −33 . 2dB ± SE 0 . 9 blindsight negative , t = 0 . 3 , p = 0 . 8 , df = 12 ) . Outside the scanner , two behavioural experiments were performed: ( 1 ) 2AFC temporal detection and ( 2 ) 2AFC direction discrimination ( S1 Fig ) . The experiments were conducted on the same day as scanning , using a 60-Hz CRT monitor at a viewing distance of 68 cm . Throughout behavioural experiments , participants were asked to maintain fixation , with the investigator observing this in real time using an Eyelink 1000 Eye Tracker ( SR Research Limited , Ontario , Canada ) . Anyone making even a small eye movement into their damaged hemifield was given specific instruction not to do so , and it was explained that these data would have to be discarded . At the start of the experiment , an identical , static test stimulus was used to confirm that patients were unable to see the stimulus at its selected size and location in the visual field . This was done using a predicted aperture size and locus based upon prior perimetry results . Stimulus location had to be restricted to the boundary of the fMRI display , which subtended 23° horizontally and 13° vertically . This influenced whether a 5° or 8° diameter stimulus was chosen , as the stimulus had to stay inside the ‘blind’ field while remaining on screen . The stimulus of choice was an 8° diameter aperture , but if this was not possible , the stimulus was reduced to 5° diameter . If the criteria were unachievable using either stimulus size , the patient was excluded from the study . If the patient was able to see any part of the test stimulus whilst fixating on the central cross , the aperture was repositioned 0 . 5° deeper into the scotoma ( according to the Perimetry report ) until the patient could no longer see any part of the stimulus at all . Any trials with eye position more than 1° from fixation were excluded from analysis . The presence or absence of residual visual function ( blindsight ) was determined according to patients’ ability to detect stimuli above chance , i . e . , Experiment 1 . Specifically , this was defined as achieving either an average score or a score for individual conditions that was significantly above chance , using a statistical threshold of p < 0 . 05 and a cumulative binomial distribution . This was an identical method to our previous work , except that the stimulus was moving dots rather than a drifting Gabor [33] . We selected stimulus location based upon perimetry results , as detailed above . Necessarily , this meant that all patients showed the same abnormal visual performance for their test locations . Aside from demonstrating the same visual sensitivity on perimetry , stimuli in blindsight-negative patients were no deeper into the visual field than they were for blindsight-positive patients , suggesting this was not a critical factor for the difference in behavioural performance . Previous blindsight studies have employed a variety of visual stimuli ( moving dots , gratings , moving bars , high luminance targets [31 , 37 , 63 , 69] ) and a number of different techniques for assessment , including 2AFC , indirect behavioural performance , saccadic eye movements , and navigational performance ( e . g . , [5 , 70–72] ) . It is also common to target only one retinal location in blindsight testing [31 , 37 , 63 , 69 , 70] . The critical point for the definition of blindsight is that patients show significant performance despite absent visual capacity in the targeted region of the visual field . We ensured that a conservative threshold was used to target truly ‘blind’ regions of scotoma , which we demonstrated to be no different in patients with or without blindsight . Since we were particularly interested in the role of hMT+ , we assessed whether a moving stimulus could be detected without awareness for our definition of blindsight . Using these criteria , eight patients were categorized as ‘blindsight positive’ , as they could detect the stimulus inside their blind hemifield significantly above chance ( P2 , P3 , P5 , P8 , P10 , P11 , P13 , P14 ) . Of these individuals , four could also discriminate motion direction above chance ( Experiment 2; P5 , P8 , P10 , P14 ) . With regard to subjective awareness , only two patients reported any awareness of the stimuli during the experiment ( P3 and P10 ) . Both had been completely unaware of static dots in the pre-experiment assessment , in which static dots were positioned at the same coordinates , without fast onset/offset . For moving stimuli , P10 reported knowing that something was there , but was unable to distinguish what it was . P3 also could not describe what she saw , suggesting she had been looking at ‘streaks or shadows’ . Both patients were at ceiling on the detection task , but P3 remained at chance on the direction discrimination task . Of note , five of the patients ( P3 , P8 , P10 , P11 , P13 ) took part in a previous fMRI study [33] , in which they also demonstrated significant blindsight performance for detection of a drifting Gabor . Eye movements were defined as a change in fixation towards the scotoma of 1 degree or more . This would capture all eye movements irrespective of their type , i . e . , saccadic , slow drift , nystagmus . The threshold of 1 degree ensured that stimuli could never be directly fixated but would always remain inside the scotoma . Although microsaccades were possible , these would not bring the visual stimulus into the seeing portion of the visual field . This methodology was the same as previous work ( Fig 2B in [33] ) , in which we also provided examples of successfully identified saccades . Seven trials were removed from analysis in Experiment 1 and 2 trials from Experiment 2 due to eye movements of more than 1 degree towards the stimulus calculated from retrospective eye tracker data analysis . At the time of the experiment , a further 6 trials were flagged for exclusion in Experiment 1 and 4 in Experiment 2 due to real-time observation of the experimenter or feedback from the patient . In total , this accounted for 0 . 93% of trials in Experiment 1 and 0 . 54% of trials in Experiment 2 that were excluded from analysis due to inappropriate eye position . The same stimuli were viewed during fMRI , presented separately to each hemifield . Stimuli during scanning were presented on a 1 , 280 × 1 , 040 resolution monitor at the back of the MRI scanner bore . Participants viewed stimuli via a double mirror mounted on the head coil . When in position , the screen subtended a visual angle of 23° × 13° . The same 5 speed levels were presented separately to each hemifield , representing a 10-condition block design ( S2B Fig ) . For each block , the aperture of moving or stationary black dots appeared for 16 s . Direction coherence was 100% , and dots moved at a constant speed . Angle of drift changed at random every two seconds from a choice of 8 directions . A 10-s rest period followed each block . Throughout all experiments , participants performed a task to maintain fixation by pressing a button every time a central fixation cross changed colour from black to red ( S2B Fig ) . Colour changes occurred at random , lasting 300 ms in duration , and participants were instructed at the start to try not to miss any red crosses . It was emphasised that they must try to maintain fixation throughout and avoid moving their eyes around the screen . An EyeLink 1000 eye tracker ( SR Research Limited , Ontario , Canada ) was again used to confirm central fixation by recording eye position ( see section fMRI eye tracking ) . Scanning took place using a 3T Siemens Verio MRI scanner at the Functional Magnetic Resonance Imaging Centre of the Brain ( FMRIB , University of Oxford ) . At the start of each sequence , magnetisation was allowed to reach a steady state by discarding the first five volumes , an automated feature of the scanner . T2*-weighted EPI volumes covered 34 sequential 3-mm slices ( repetition time , TR 2000 ms; echo time , TE 30 ms ) with three runs , each lasting 260 s . In a single session lasting 13 . 2 min , 396 functional volumes were acquired . For one participant ( P4 ) we collected one additional session of fMRI data . For three patients , we collected four runs of data ( i . e . , 526 volumes , P1 , P2 , P3 ) . For one patient ( P7 ) and one control , we only collected two runs of data , i . e . , 266 volumes ( 8 . 9 min ) . We also acquired a high-resolution ( 1 mm3 ) whole-head T1-weighted MPRAGE anatomical image ( TE 4 . 68 ms; TR 2040 ms; flip angle , 8° ) and a field map ( TE1 , 5 . 19 ms; TE2 , 7 . 65 ms; 2 mm3 ) for each participant . fMRI preprocessing and statistical analyses were carried out using tools from FMRIB's Software Library ( FSL , www . fmrib . ox . ac . uk/fsl ) . Non-brain tissue was excluded from analysis using the Brain Extraction Tool ( BET ) [73] , motion correction was carried out using MCFLIRT [74] , and images were corrected for distortion using field maps . For cortical ROIs and group contrast maps , spatial smoothing used a Gaussian kernel of FWHM 5 mm , and high-pass temporal filtering ( Gaussian-weighted least-squares straight line fitting , with sigma = 13 . 0 s ) was applied . For all subcortical ROI analyses , no spatial smoothing was applied to ensure that signals were not contaminated with adjacent structures . Functional images were registered to high-resolution structural scans using FLIRT [75] and to a standard Montreal Neurological Institute ( MNI ) brain template using FLIRT and FNIRT [76] . This enabled us to transform anatomical and probabilistic regions of interest into functional space for analysis ( see S3 Fig for an illustration of subcortical ROIs in functional , structural , and standard space for each participant ) . Eye movements during fMRI can be a legitimate concern when considering results for visual stimulation inside a scotoma . In this study , three main lines of evidence suggest that this was not a problem and could not have accounted for the results . First , concurrent eye movement data was collected on most patients using an eye tracker positioned at the base of the MRI bore ( n = 10 ) . All of these patients underwent successful eye-tracker calibration , with accurate data throughout fMRI runs . For this group of patients , the mean number of eye movements was 5 . 8 ± 3 . 5 SEM , defined as a movement of 1 . 5 degrees or more towards the scotoma . This accounted for <0 . 3% of the scan duration , suggesting that any effects on the results are likely to be negligible . To confirm this , when scanner volumes corresponding to eye movements were regressed out of analyses , the results remained unchanged ( r > 0 . 99 ) . For the patients without eye movement data , there had been difficulty either with calibration due to their dense field loss or with visualisation due to the presence of corrective acuity lenses . In those situations , direct visualisation was used via video recording of the pupil to observe any overt eye movements during the experiment . Second , participants performed over 90% on a concurrent behavioural task that required fixation throughout the experiment ( S2C Fig ) . Brief colour changes of the fixation cross ( 300-ms duration ) occurred at frequent and random intervals , and participants were given a window of 1 s to press a button connected to the stimulus computer via a parallel port , being specifically instructed not to miss any red crosses or move their eyes around the screen . In addition , before the fMRI scan , all participants took part in behavioural testing lasting at least 60 min , focussed on their damaged region of vision . Participants became very experienced at maintaining fixation during this assessment . hMT+ masks were derived from probabilistic maps ( Jülich atlas implemented in FSL ) [77 , 78] . These were transformed into functional space for patients and controls to ensure consistency between participant groups . V1 masks in controls and in the undamaged hemisphere of patients were functionally defined so that they corresponded to stimulated regions of calcarine cortex . In native space , average hMT+ ROI volume was 94 . 8 ± 35 . 2 SD voxels in patients and 100 . 9 ± 42 . 0 SD voxels in controls ( t = 0 . 5 , p = 0 . 6 , df = 42 ) . Average V1 ROI volume was 16 . 2 ± 7 . 5 SD voxels in patients ( undamaged hemisphere ) and 24 . 4 ± 7 . 1 SD voxels in controls ( averaged across hemispheres ) , the small volume reflective of the small 5°- or 8°-diameter stimulus used . For the LGN and SC , binary masks were created by manual inspection and drawing over the anatomical T1-weighted images [79] , using a radiological brain atlas to aid identification of landmarks ( See S3 Fig for masks in all patients ) . The average LGN volume in patients measured 248 mm3 in the right and 246 mm3 in the left . In controls , average LGN volume was 240 mm3 in the right and 239 mm3 in the left . The average SC volume in patients was 195mm3 in the left and 177 mm3 in the right . For the ventral pulvinar , binary masks were created in MNI152 standard space according to the description of Arcaro and colleagues [50] . It was possible to visualise the nucleus as a region of low T1 intensity relative to surrounding tissue in the posterior most part of the thalamus . Masks were transformed to anatomical and functional space for each participant and were manually inspected to ensure accuracy ( S3 Fig ) . Average ventral pulvinar volume was 385 mm3 in the left and 365 mm3 in the right . There were no significant differences in ROI volume between blindsight-positive and -negative patients . Lesion masks were drawn manually in structural space ( S4 Fig ) . For S2D and S2E Fig , these were nonlinearly transformed to standard space and binarised before being summed . To assess whether lesions encroached upon hMT+ and/or its surrounding white matter , we created subject-specific cuboidal ROIs that were centred on the ‘centre of gravity’ of the ipsilesional hMT+ ROI in structural space . The isotropic cubes measured 40 × 40 × 40 mm3 , thus containing 64 , 000 1-mm3 voxels ( see S5A and S5B Fig for examples in P6 and P12 ) . We superimposed the binarized lesion mask over the cuboidal mask and counted the number of voxels that overlapped ( red voxels in S5B Fig ) . The voxel counts are shown in the table in S5 Fig . All graphs , signal change calculations , and correlation statistics were calculated using data from participants’ native space . For region of interest analysis , each experimental condition ( e . g . , left hemifield , 8°/s speed ) was entered into the general linear model as a separate explanatory variable and was contrasted against the baseline fixation task to generate contrast of parameter estimates ( COPEs ) for each condition in every voxel . Signal change was then extracted from regions of interest within functional-space for each individual . The percentage of signal change was calculated by scaling the COPE by the peak-peak height of the regressor and dividing by the mean over time . These measures were averaged across participants to generate group plots for signal change as a function of the condition under investigation and were used in all correlation and regression analyses . For whole-brain group analyses ( Figs 4 and 5 ) , it was necessary to align patient brains to a uniform pathological template with lesions located in the same ‘left’ hemisphere corresponding to a ‘right-sided’ visual field deficit . This required that the structural and functional images of three patients ( P3 , P5 , P12 ) be flipped in the horizontal plane . All activation coordinates and images were in MNI space , with beta values displayed on mean structural images for the group transformed to standard space . For the whole time series analyses , a value for residual BOLD signal in the ROI was obtained for each volume , and this was plotted against time . This was done separately for each participant and performed in functional space . As control participants demonstrated slightly different hMT+ localization in left and right hemispheres , we decided to show hMT+ group maps for left and right V1 lesions separately . This generated a group size of n = 6 for blindsight-positive patients with left V1 lesions ( Fig 1Ai ) , n = 5 for blindsight-negative patients with left V1 lesions ( Fig 1Bi ) , n = 2 for blindsight positive-patients with right V1 lesions ( Fig 1Aii ) , and n = 1 for blindsight-negative patients with a right V1 lesion ( Fig 1Bii ) . Mixed effects analyses were used for all group analyses where n > 3 . A statistical threshold of p < 0 . 001 uncorrected was used to test for significance within V1 and extrastriate cortex , for which there were a priori hypotheses . Elsewhere , correction for multiple comparisons was made using a cluster threshold of p < 0 . 05 unless otherwise stated . Statistical tests to quantify differences in functional activity and co-activation between ROIs or participant groups were implemented in Excel or MATLAB . For overall hMT+ percent BOLD responses ( Fig 1 , bar charts ) , activity was averaged across all five motion speeds ( 0–32°/s ) . Two statistical analyses were then performed: ( i ) mean activity comparing sighted and blind hemifield ( and hemisphere ) used a paired t test and ( ii ) mean activity compared to baseline using a one-sample t test versus zero . A two-way ANOVA was also used to assess the effect of participant group and speed on blind hemifield responses ( right hemifield in controls ) and separately on sighted hemifield responses ( left hemifield in controls ) . In fMRI time series correlations , it is known that task conditions can influence intrinsic temporal correlations ( e . g . , see [12 , 17] ) . To ensure that correlations only reflected resting block activity , we used the residuals timeseries for ROIs once stimulus responses had been regressed out . This allowed us to determine resting ROI1 versus ROI2 correlation analyses for each participant . For correlation analyses , two main statistical methods were used . For correlations between participant groups , a Pearson correlation coefficient was derived from mean activity in ROI1 versus ROI2 at each level of speed , i . e . , n = 5 . For correlations between ROIs within participant groups , a Pearson correlation coefficient was determined separately for each participant ( activity in ROI1 versus ROI2 , at each level of speed ) . We then calculated weighted averages of r coefficients for the group , using a Fischer transformation to approximate correlations to a normally distributed measure ( Figs 2 and 3 ) . A significant effect of group was determined by performing a one-way ANOVA . Pairwise comparisons between participant groups or ROIs were then calculated using post-hoc t tests . One-sample t tests were used to compare r coefficients to zero . Whenever activity was compared in the same participant , a paired correlation analysis was performed . For each participant , the raw signal time series for the seed ROI was entered into the model as an explanatory variable . This was applied to the filtered and motion corrected whole brain timecourse . Stimulus conditions were also entered as regressors so that the model would better describe the data . Since the model was identical to the ROI time series , the parameter estimate was always 1 in the seed region . Any participant with voxels outside the seed region with a parameter estimate >1 . 5 were excluded from analysis , as these results were likely to be driven by noise . This was only a problem when using subcortical structures as seeds , as these are small regions with relatively weak signal that are more susceptible to artefact . This led to the exclusion of three participants from subcortical seed analyses ( both hemispheres ) ; two were blindsight positive ( P5 , P10 ) , and one was blindsight negative ( P7 ) . Corroborating this , the raw LGN signal range in those three participants was significantly greater than the other 11 patients ( 66 . 6 ± SD 10 . 5 versus 39 . 6 ± SD 6 . 07 , t = 2 . 2 , p = 0 . 04 ) and controls ( 30 . 6 ± 1 . 9 , t = 5 . 2 , p < 0 . 001 , df = 12 ) . To generate seed region correlation maps ( Figs 4 and 5 ) , the COPEs for each included participant were entered into a higher-level mixed effects analysis , and output parameter estimate ( beta ) maps were used to represent seed region correlation maps . This was performed separately for each participant group and for each seed region . The resulting maps were not intended to determine statistical significance but to allow visual inspection of the results from the separate groups . Visualisation thresholds were based upon control participants , and an optimal cutoff was used to display correlations in either V1 , visual subcortex , and/or hMT+ without excess background noise . The same thresholds were applied to all participant groups . For LGN-oriented maps ( Fig 5Aiv–5Avi and 5Biv–5Bvi ) , although not shown , no other subcortical regions showed equivalent beta levels , although ventral pulvinar was co-activated in both hemispheres of all participant groups when using a reduced threshold of 0 . 32 .
|
When the primary visual cortex ( V1 ) is damaged in one hemisphere , we lose the ability to see one half of the world around us . Clinical tests show that in this blind region of vision , we cannot see even the brightest flashes of light . However , many years of research have shown that individuals who are blind in this way may still respond to certain images in the ‘blind’ area of vision , even though they are often unable to describe what they ‘see’ and may be unaware of seeing anything at all . This is called blindsight , and researchers are trying to understand the pathways underlying this phenomenon . A recent study mapped a physical pathway of connections in the brain that could account for blindsight in humans . However , the functional nature of this pathway has never been shown . In this study , we assess a group of patients with damage to V1 , some of whom demonstrate blindsight and some of whom do not . We compare neural responses and functional connectivity and show that a functional connection in this pathway is critical for blindsight . We also reveal new insights into how speed and motion are likely to be processed in the healthy brain .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"diagnostic",
"radiology",
"pathology",
"and",
"laboratory",
"medicine",
"functional",
"magnetic",
"resonance",
"imaging",
"brain",
"social",
"sciences",
"neuroscience",
"cerebral",
"hemispheres",
"left",
"hemisphere",
"magnetic",
"resonance",
"imaging",
"signs",
"and",
"symptoms",
"scotoma",
"brain",
"mapping",
"vision",
"neuroimaging",
"blindness",
"research",
"and",
"analysis",
"methods",
"sensory",
"physiology",
"imaging",
"techniques",
"behavior",
"lesions",
"visual",
"system",
"psychology",
"radiology",
"and",
"imaging",
"eye",
"movements",
"diagnostic",
"medicine",
"visual",
"impairments",
"anatomy",
"physiology",
"ophthalmology",
"biology",
"and",
"life",
"sciences",
"sensory",
"systems",
"sensory",
"perception"
] |
2018
|
Blindsight relies on a functional connection between hMT+ and the lateral geniculate nucleus, not the pulvinar
|
Eumycetoma is a progressive and destructive chronic granulomatous subcutaneous inflammatory disease caused by certain fungi , the most common being Madurella mycetomatis . The host defence mechanisms against fungi usually range from an early non-specific immune response to activation and induction of specific adaptive immune responses by the production of Th-1 and Th-2 cytokines . The aim of this study is to determine the levels of Th-1 and Th-2 cytokines in patients infected with Madurella mycetomatis , and the association between their levels and disease prognosis . This is a descriptive cross-sectional study conducted at the Mycetoma Research Centre , University of Khartoum , Sudan , where 70 patients with confirmed M . mycetomatis eumycetoma were enrolled; 35 with , and 35 without surgical excision . 70 healthy individuals from mycetoma endemic areas were selected as controls . The levels of serum cytokines were determined by cytometric bead array technique . Significantly higher levels of the Th-1 cytokines ( IFN-γ , TNF-α , IL-1β and IL-2 ) were recorded in patients treated with surgical excision , compared to those treated without surgical excision . In contrast , the Th-2 cytokines ( IL-4 , IL-5 , IL-6 and IL-10 ) were significantly lower in patients treated with surgical excision compared to those treated without surgical excision . In conclusion , the results of this study suggest that cell-mediated immunity can have a role to play in the pathogenesis of eumycetoma .
Mycetoma is a chronic subcutaneous infection caused by certain bacteria ( actinomycetoma ) or fungi ( eumycetoma ) [1] . It is characterised by a slow progressive infection and a granulomatous inflammatory response that can result in severe soft tissue and muscle damage along with destruction of the underlying bone [1 , 2] . Mycetoma is endemic in tropical and subtropical regions; however , it has been reported globally . Eumycetoma in Sudan , is predominately caused by the fungus Madurella mycetomatis [2] . The disease is characterised by extensive subcutaneous masses , usually with multiple draining sinuses and fungal grains [1] . Mycetoma disease has significant negative medical health and socio-economic impacts on patients and communities , affects individuals of all ages , but is more frequently seen in adults who work outdoors . The host defence mechanisms against fungi usually range from germline encoded immunity which present early in the evolution of microorganisms , to highly specialised and specific adaptive mechanisms that are induced by infection and disease . The innate response to fungi serves two main purposes; a direct antifungal effector activity and activation or induction of specific adaptive immune responses . In general , the direct antifungal effector activity mediates non-specific elimination of pathogens through either a phagocytic process with intracellular killing of internalised pathogens or through the secretion of microbiocidal compounds against undigested fungal molecules . The activation and induction of the specific adaptive immune responses is accomplished by the production of pro-inflammatory mediators , including chemokines and cytokines , providing co-stimulatory signals to naive T cells , as well as antigen uptake and presentation to CD4+ and CD8+ T cells [3 , 4] . Many individuals in mycetoma endemic areas are exposed to the causative aetiological agents , but only few develop the disease . This may suggest variable responses of the host immune system towards the invading agent . In this respect , the role of the innate immunity in host resistance to mycetoma infection has been studied in vitro and in animal models , but few studies have been performed in humans . T cell–mediated immune response to eumycetoma fungi in humans was studied by Mahgoub and associates who suggest that patients with eumycetoma have a weak cell-mediated response as determined by skin reaction to dinitrochlorobenzene [5] . Decreased lymphocyte proliferative response to phytohemagglutinin in those patients was also reported . However , no evidence was provided to confirm whether this is a primary immune deficiency or a secondary response to a severe infection . In addition , the same study showed high levels of IgA and IgM and low levels of IgG antibodies in mycetoma patients [5] . In actinomycetoma , Gonzalez-Ochoa and Baranda [6] found that patients with severe lesions and extensive tissue destruction displayed a weak skin reaction to some pathogenic bacteria polysaccharides such as , Nocardia brasiliensis [6] . However , it was not clear whether this represented a T-helper-1 ( Th-1 ) or T-helper-2 ( Th-2 ) response . To date there has been limited data on the immune response to mycetoma infection and how patients can modulate their response against M . mycetomatis . With this background , the present study aims to determine the Th-1 and Th-2 cytokines response of patients infected with M . mycetomatis and to find out the association between the measured Th-1 and Th-2 cytokine levels and the disease prognosis and outcome .
This descriptive cross-sectional hospital based study was conducted at the Mycetoma Research Centre , University of Khartoum , Khartoum , Sudan . In this study 140 individuals were enrolled; 49 ( 35% ) were females and 91 ( 65% ) were males ( Table 1 ) , with an overall median age of 25 years ( range 12–70 years ) . 70 patients with confirmed mycetoma infection due to Madurella mycetomatis were recruited . The study population was divided into three groups; group I: healthy controls ( n = 70; median age 25 years ( range 12 to 70 years ) ) , matched for sex , age and locality with the patients group . Group II: mycetoma patients without surgical excision ( n = 35 patients; median age 25 ( range 13 to 70 years ) ) , these patients were not treated with surgical excision and were under medical treatment ( 200 mg bd Itraconazole or 400 mg bd ketoconazole ) . Group III: mycetoma patients who underwent surgical excision ( n = 35 patients; median age 25 years ( range 12 to 70 years ) and medical treatment ( 200 mg bd Itraconazole ) . One hundred μl of blood were collected on filter paper ( Whatman qualitative filter paper , Grade 1 , circles , diam . 42 . 5 mm from SIGMA-ALORICH , KSA ) for cytokine’s determination . The use of filter paper dried whole blood spots ( DBS ) for specimen collection was preferred to facilitate collection , storage and transportation of specimens in addition to being recommended by the World Health Organization ( WHO ) and also used in several previous studies [7–9] . A hole puncher with a diameter of 6 mm was used for cutting out discs from the filter paper in the middle of the blood spot , where the blood was assumed to be evenly spread . The discs were put in 10 ml tubes and 500 μl of PBS containing 0 . 05% Tween and 0 . 5% BSA was added . The discs were then incubated for 2 hrs at room temperature on a shaker . Finally , after vortexing the samples for 30 seconds , the supernatants ( eluted serum ) were collected with a Pasteur pipette and aliquoted in new 1 . 5 ml cryo tubes and stored at −20°C until analysis . The extract corresponds to a serum dilution of ~1:100 . This method was modified from a previous report by Mercader and colleagues [8] . Measurements of cytokines were performed in sera by flow cytometry using Cytometric Bead Array ( CBA ) technology , as detailed by Cook and associates [10] . Human Inflammation CBA kit ( BD Biosciences , San Jose , CA ) was used to quantitatively measure IFN-γ , TNF-α , IL-1β , IL-2 , IL-4 , IL-5 , IL-6 , IL-10 , and IL-13 levels . The sensitivity of Human Inflammation CBA was comparable to conventional ELISA [11] . Samples were analysed using a BD FACSCalibur flow cytometer ( BD Biosciences , San Jose , CA ) , according to the manufacturer’s instructions . The data was managed by SPSS statistics software version 23 for Windows ( IBM , SPSS statistics ) . The one-way analysis of variance ( ANOVA ) and Tukey’s test for post hoc analysis were used to compare mean levels of cytokines between various study groups . The difference in cytokine levels across groups was analysed using ANOVA test ( Table 2 ) . Linear regression models were used to predict each cytokine level ( Table 3 ) . Unstandardised coefficient ( B ) regression is the determination of the statistical relationship between two or more variables [12] . B analysis was adjusted for each cytokine according to gender ( Female = 0 and male = 1 ) , medical treatments ( Itraconazole = 0 and Ketoconazole = 1 ) , size of mass ( >10 cm = 1 ) , presence of grains ( No = 0 and Yes = 1 ) and age , as independent variables . This study was approved by the Ethics Committee of Soba University Hospital , Khartoum , Sudan . Written informed consent was obtained from the participants prior to their enrolment in the study . Informed consent was also obtained from children and their guardians before participation . The work described here was performed in accordance with the Declaration of Helsinki [13] .
A higher proportion of mycetoma patients were males ( 80% ) compared with females ( 20% ) . Combined ( both with and without surgical intervention ) males and females among mycetoma patients groups were 56/70 and 14/70 , respectively ( p <0 . 001; Table 1 ) . Patients with mycetoma received various antifungal drugs , which were used in combination with or without surgical excision . Of the 70 individuals who received oral medication in this study , 46 patients ( 66% ) received Itraconazole . Out of the patients who were treated with Itraconazole , Eleven patients ( 24% ) were treated without surgical excision and 35 patients ( 76% ) were surgically treated along with Itraconazole 200 mg bd . Twenty four patients ( 34% ) received Ketoconazole 400 mg bd [p value <0 . 001 and 95% confidence interval 95%CI; ( 0 . 55 to 0 . 80 ) ] ( Table 1 ) . Ketoconazole 400 mg bd was only used among patients without surgical excision and not following surgery , whereas Itraconazole 200 mg bd was the only choice postoperatively [p value <0 . 001 and 95% CI; ( 0 . 58 to 0 . 93 ) ] . The proportion of lesions that were more than 10 cm in diameter were significantly higher in the surgically treated group compared to the non-surgically treated patients [p value = 0 . 037 and 95% CI; ( -0 . 58 to 0 . 13 ) ] ( Table 1 ) . Patients with mycetoma infection had significantly higher cytokine levels including IFN-γ , TNF-α , IL-2 , IL-4 , IL-5 , IL-6 , IL-10 and IL-13 , compared to the control group ( overall p value for each cytokine <0 . 001 ) ( Table 2 ) . In contrast; no significant difference was observed in the levels of IL-1β between the study groups ( overall p value = 0 . 913 ) ( Table 2 ) . Linear regression analysis showed significantly higher levels of Th-1 cytokines ( IFN-γ , TNF-α , IL-1β and IL-2 ) among mycetoma patients treated with surgical excision than in those treated without surgical intervention . Unadjusted B ( 95% CI ) for: IFN-γ = [5 . 64; 95% CI ( 1 . 33 to 9 . 96 ) , p value = 0 . 011] . For TNF-α = [14 . 58; 95% CI ( 11 . 56 to 17 . 60 ) , p value <0 . 001] . For IL-1β = [-0 . 36; 95% CI ( -0 . 67 to -0 . 05 ) , p value = 0 . 022] . For IL-2 = [7 . 55; 95% CI ( 5 . 61 to 9 . 50 ) , p value <0 . 001] ( Table 3 ) . When B was adjusted for gender , medical treatment , size of lesions and the presence of grains; similar statistical analysis indicated significantly higher levels of Th-1 cytokines ( IFN-γ , TNF-α , IL-1β and IL-2 ) among mycetoma patients treated with surgical excision than in those treated without surgical excision . Adjusted B ( 95% CI ) for: IFN-γ = [6 . 62; 95% CI ( 1 . 42 to 11 . 81 ) , p value = 0 . 017] . For TNF-α = [12 . 69; 95% CI ( 9 . 94 to 16 . 32 ) , p value <0 . 001] . For IL-1β = [-0 . 75; 95% CI ( -1 . 13 to -0 . 37 ) , p value <0 . 001] . For IL-2 = [6 . 59; 95% CI ( 3 . 91 to 9 . 28 ) , p value <0 . 001] ( Table 3 ) . In contrast , a similar linear regression analysis model for Th-2 cytokines showed significantly lower levels of Th-2 cytokines ( IL-4 , IL-5 , IL-6 and IL-10 ) among mycetoma patients treated with surgical excision , compared to those treated without surgical excision . Unadjusted B ( 95% CI ) for: IL-4 = [-2 . 57; 95% CI ( -3 . 14 to -2 . 0 ) , p value <0 . 001] . For IL-5 = [-2 . 08; 95% CI ( -2 . 54 to -1 . 62 ) , p value <0 . 001] . For IL-6 = [-10 . 09; 95% CI ( -13 . 68 to -6 . 51 ) , p value <0 . 001] . For IL-10 = [-5 . 33; 95% CI ( -7 . 79 to -2 . 87 ) , p value <0 . 001] ( Table 3 ) . When B was adjusted for gender , medical treatment , size of lesions and presence of grains , a similar statistical analysis model showed significantly lower levels of Th-2 cytokines ( IL-4 , IL-5 , IL-6 and IL-10 ) among mycetoma patients treated with surgical excision compared to those treated without surgical excision ( Table 3 ) . Adjusted B ( 95% CI ) for: IL-4 = [-2 . 82; 95% CI ( -3 . 65to -1 . 99 ) , p value <0 . 001] . For IL-5 = [-2 . 38; 95% CI ( -3 . 04 to -1 . 72 ) , p value <0 . 001] . For IL-6 = [-7 . 66; 95% CI ( -12 . 88 to -2 . 44 ) , p value = 0 . 005] . For IL-10 = [-3 . 58; 95% CI ( -7 . 16 to -0 . 01 ) , p value = 0 . 05] ( Table 3 ) .
It is known that fungi release antigens ( Ag ) on the skin surface , and the antigens that penetrate the skin are subsequently captured by an antigen-presenting cell ( APC ) such as dendritic cells ( DCs ) [14] . Fungal antigens can also play an important role in the DCs maturation . Furthermore , production of inflammatory cytokines such as IFN-γ and TNF-α by other innate cells such as natural killer cells ( NK ) further enhance the activation of microbiocidal functions of phagocytic cells as well as maturation of DCs [15] . In the present study , Th-1 cytokines ( IFN-γ , TNF-α , and IL-2 ) were found to be significantly higher in mycetoma patients than in controls . Besides , the levels of Th-1 ( IFN-γ , TNF-α , IL-1β and IL-2 ) were significantly higher in mycetoma patients treated with surgical excision compared to those who were only medically treated . These findings go a long way to explain the earlier findings of van de Sande and associates [16] , that neutrophils are attracted to the site of infection by mycetoma antigen , secrete TNF-α and IFN-γ cytokines in the presence of IL-17 [17] . Interestingly , in a previous study , Cassatella and colleagues suggested that neutrophils are multipurpose cells which play many roles , not only in inflammatory progressions but also in immune and antitumor processes [18] . The same group had also added that , IFN-γ activated neutrophils release biologically active TNF-α related apoptosis-inducing ligand ( TRAIL/APO2 ligand ) , a molecule that exerts selective apoptotic activities towards tumours [18] . Additionally , Elagab and associates , showed that , the peripheral blood mononuclear cells ( PBMC ) of mycetoma patients react differently to M . mycetomatis antigens than healthy controls [19] . In general , when PBMCs produce IFN-γ upon stimulation with the antigen , no production of IL-10 was detected [19] . There is also no significant differences between the cytokines TNF-α and TGF-β levels in patients and controls [19] . The discrepancy between Elagab’s findings [19] and our findings may be explained by the differences in the study design . IL-1 is an essential host defence cytokine against a broad range of pathogens , ranging from bacteria to parasites and fungi [20] . IL-1β is primarily produced by innate immune cells such as monocytes , macrophages and dendritic cells upon activation , and is also an important cytokine for the control of fungal infection [21] . It is also an important proinflammatory mediator whose production is controlled by multiprotein complexes called inflammasomes [22 , 23] . Although IL-1β plays an active role in containing infection caused by different fungi , its role in controlling fungal infections remains unclear [24] . The results of the current study has shown that higher levels of IL-1β cytokine are strongly associated with mycetoma patients treated with surgical excision , compared to those treated without surgical intervention . It is of interest to note that , IL-1β can play a crucial role in the activation of complement protein-3 ( CR3 ) , dectin-1 as well as caspase-8 in coordinating cell death and inflammasome responses to β-glucans [25] . Our findings led us to suggest that the observed higher levels of IL-1β cytokine play an important role in reducing the risk of M . mycetomatis infection . However , more studies are needed to confirm farther this observation . As mentioned earlier cytokine IL-2 exerts critical functions during immune homeostasis via its effects on Treg cells , and by optimising the effector lymphocyte responses of both T-cells and B-cells . In addition , IL-2 receptors ( IL-2R ) were shown to be present on human neutrophils , and that IL-2-neutrophil interactions are believed to be important in both tumour rejection and increased susceptibility to bacterial infections [26 , 27] . It is relevant to add that a previous study on mycetoma patients from an endemic area [16] , demonstrated that neutrophils are attracted to the site of infection by the mycetoma antigen . In the current study IL-2 levels were significantly higher in mycetoma patients compared to controls . In addition , IL-2 cytokine levels were elevated significantly in mycetoma patients treated with surgical excision , compared to those treated without surgical intervention . We take this finding to indicate that , IL-2 cytokine plays a major role in the pathogenesis of mycetoma infection . This novel finding on an association of IL-2 and neutrophils should pave the way to new avenues of research on IL-2-neutrophil interactions to better understand the response of patients to mycetoma infection . The cytokines IL-4 , IL-5 , IL-13 and GM-CSF are produced by T-helper-2 cells at the site of inflammation but also they have important functions in haematopoiesis . These cytokines , individually or collectively along with chemokines such as CCL11 , play a major role in coordinating the maturation and mobilisation of leukocytes ( Monocytes/Macrophages and Neutrophils ) and mast cell progenitors , ensuring the continued supply of leukocytes to the site of the inflammation [28 , 29] . In present study , the in vivo effect of M . mycetomatis infection on the production of Th-2 cytokines ( IL-4 , IL-5 , IL-6 and IL-10 ) was clearly reflected by the , significantly higher levels of Th-2 cytokines in mycetoma patients compared to controls . Moreover , lower levels of Th-2 cytokines ( IL-4 , IL-5 , IL-6 and IL-10 ) were significantly associated with mycetoma patients treated with surgical excision , compared to those treated without surgical intervention . This finding is in line with the earlier hypothesis that Th-2 cytokines play an important role in the activation of the humoral immune response [28 , 29] . It is well stablished that the type of cell-mediated immunity ( CMI ) is critical in determining resistance or susceptibility to fungal infection . In general , Th1-type CMI is required for the clearance of fungal infections , while Th2 immunity usually enhances the susceptibility to infection and allergic responses [30] . Additionally , Th-1 cells are concerned mainly with production of cytokines such as IFN-γ , and promote CMI and phagocyte activation , while in contrast , Th-2 cells predominantly produce cytokines such as IL-4 and IL-5 and tend to promote antibody production [30–32] . Besides , IL-4 and IL-5 cytokines can play an important role in the activation of B-cells to differentiate to plasma cells that secrete IgM antibody and also generate memory B cells [33] . A previous similar study found elevated levels of IgM antibody in mycetoma patients [5] . Besides , another study on immune responses against mycetoma Sudanese patients , demonstrated the presence of immunoglobulins G , M and complement on the surface of the grains and on the filaments inside the grains of mycetoma lesions [34] . Also , both neutrophils and macrophages were recruited into the lesion by complement and were involved in the fragmentation of the grains . The cytokines profile in the lesion and regional lymph nodes was of a dominant Th-2 pattern ( IL-10 and IL-4 ) [34] , and these elevated levels of Th-2 cytokines in mycetoma patients may trigger the increased production of IgG , IgM and complement . The significance of this phenomenon needs further investigations . We noted with great interest higher levels ofTh-1 cytokines ( IFN-γ , TNF-α , IL-1β and IL-2 ) in mycetoma patients treated with surgical excision than in those patients treated without surgical intervention . However , in contrast the Th-2 cytokines ( IL-4 , IL-5 , IL-6 and IL-10 ) were significantly lower in patients treated with surgical excision compared to those treated without surgical intervention . These results suggest that , the defence against the fungus M . mycetomatis is based on the adaptive effector phase and the duration of the infection as well as the size of the mycetoma mass and presence of grains . The effects of CMI can also play a critical role in reducing the risk of localised infection in mycetoma patients treated with surgical excision compared to those treated without surgical intervention . The essential role of the CMI response is to destroy the fungi and produce an immuno-protective status against infection . At this moment the exact explanation of this finding is not clear and requires further investigation in mycetoma patients .
|
Madurella mycetomatis is the most common causative agent for eumycetoma , which is a progressive and destructive subcutaneous inflammatory disease . It is a neglected tropical disease affecting the population in poor and remote endemic tropical and subtropical areas . Currently , the susceptibility and resistance to mycetoma are not well defined , and many factors can be incriminated , including immunological , genetic , or environmental ones . The current descriptive cross-sectional study was conducted to determine the Th-1 and Th-2 cytokine levels among 70 patients with Madurella mycetomatis eumycetoma and 70 healthy controls . It aimed to find out the association between the disease prognosis and the level of these cytokines . Significantly higher levels of the Th-1 cytokines ( IFN-γ , TNF-α , IL-1β and IL-2 ) were found in patients treated with surgical excision compared to those treated without surgical intervention . However , the Th-2 cytokines ( IL-4 , IL-5 , IL-6 and IL-10 ) were significantly lower in patients treated with surgical excision compared to those treated without surgical excision . These findings suggested that , cell-mediated immunity has a prime role in the pathogenesis of eumycetoma .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"cytokines",
"pathology",
"and",
"laboratory",
"medicine",
"mycetoma",
"immunology",
"tropical",
"diseases",
"surgical",
"and",
"invasive",
"medical",
"procedures",
"developmental",
"biology",
"fungi",
"signs",
"and",
"symptoms",
"molecular",
"development",
"neglected",
"tropical",
"diseases",
"fungal",
"diseases",
"infectious",
"diseases",
"lesions",
"immune",
"response",
"immune",
"system",
"diagnostic",
"medicine",
"surgical",
"excision",
"physiology",
"biology",
"and",
"life",
"sciences",
"organisms"
] |
2016
|
Th-1, Th-2 Cytokines Profile among Madurella mycetomatis Eumycetoma Patients
|
Fragile X syndrome ( FXS ) is a form of inherited mental retardation in humans that results from expansion of a CGG repeat in the Fmr1 gene . Recent studies suggest a role of astrocytes in neuronal development . However , the mechanisms involved in the regulation process of astrocytes from FXS remain unclear . In this study , we found that astrocytes derived from a Fragile X model , the Fmr1 knockout ( KO ) mouse which lacks FMRP expression , inhibited the proper elaboration of dendritic processes of neurons in vitro . Furthermore , astrocytic conditioned medium ( ACM ) from KO astrocytes inhibited proper dendritic growth of both wild-type ( WT ) and KO neurons . Inducing expression of FMRP by transfection of FMRP vectors in KO astrocytes restored dendritic morphology and levels of synaptic proteins . Further experiments revealed elevated levels of the neurotrophin-3 ( NT-3 ) in KO ACM and the prefrontal cortex of Fmr1 KO mice . However , the levels of nerve growth factor ( NGF ) , brain-derived neurotrophic factor ( BDNF ) , glial cell-derived neurotrophic factor ( GDNF ) , and ciliary neurotrophic factor ( CNTF ) were normal . FMRP has multiple RNA–binding motifs and is involved in translational regulation . RNA–binding protein immunoprecipitation ( RIP ) showed the NT-3 mRNA interacted with FMRP in WT astrocytes . Addition of high concentrations of exogenous NT-3 to culture medium reduced the dendrites of neurons and synaptic protein levels , whereas these measures were ameliorated by neutralizing antibody to NT-3 or knockdown of NT-3 expression in KO astrocytes through short hairpin RNAs ( shRNAs ) . Prefrontal cortex microinjection of WT astrocytes or NT-3 shRNA infected KO astrocytes rescued the deficit of trace fear memory in KO mice , concomitantly decreased the NT-3 levels in the prefrontal cortex . This study indicates that excessive NT-3 from astrocytes contributes to the abnormal neuronal dendritic development and that astrocytes could be a potential therapeutic target for FXS .
Fragile X syndrome is a form of inherited mental retardation in humans that results from expansion of a CGG repeat in the Fmr1 gene on the X chromosome [1] , [2] . This syndrome is characterized by low intelligence quotient , attention deficits , and anxiety [3]–[6] . As an mRNA binding protein , FMRP is associated with polyribosomes and involved in the translational efficiency and/or trafficking of certain mRNAs [7] . FMRP is widely expressed in the brain [8] , [9] , and its absence is expected to disrupt the synthesis and/or the subcellular localization of several proteins , which is important in long-term synaptic plasticity [10] , [11] . Fmr1 knockout ( KO ) mice serve as a model to study fragile X mental retardation [12] , [13] . Our previous study has identified FMRP as a key messenger for dopamine modulation in the forebrain and provided insights on the cellular and molecular mechanisms underlying FXS [14] . However , the cellular pathophysiology of FXS is still under discussion . Emerging evidence suggests that glia may also be involved in the development of FXS . For example , astrocytes , the major glia of the central nervous system ( CNS ) , have been shown to regulate the stability , dynamics , and maturation of dendritic spines [15] , [16] and take part in the regulation of synaptic plasticity and synaptic transmission [17] , [18] . FMRP is expressed in the astrocyte lineage during development [19] , and a lack of FMRP in astrocytes affects the dendritic morphology of neurons [20] . Evidence is steadily implicating astrocytes in synaptic maturation and elimination , suggesting that FMRP may be essential to the role of astrocytes in synaptogenesis during development [19] . However , the underlying mechanisms of astrocytes in regulating neuronal dendritic development in FXS are still unclear . In this study , we found that a lack of FMRP leads to an overexpression of neurotrophin-3 ( NT-3 ) and this in turn reduces dendritic growth in neurons . Therefore , excessive NT3 from astrocytes contributed to the dendritic developmental disorder of Fmr1 KO mice . The present study indicates an important role of normal astrocyte secretion in neuronal dendritic development and provides a potential target for FXS treatment .
FXS patients and Fmr1 KO mice have been known to have abnormal dendritic arbors with increased branch density in neurons [21] . Astrocytes are required for the efficient formation , maturation , and maintenance of synapses [17] , [18] . To explore the role of astrocytes in FXS , we isolated astrocytes from both WT and KO newborn mice with over 95% purity , as determined by the cell specific marker GFAP staining . The average area of single astrocytes was not different between the WT and KO astrocytes ( Figure 1A ) . Western blot analysis indicated that the astrocytes expressed FMRP in WT mice but not in Fmr1 KO mice ( Figure 1B ) . This result is consistent with that of previous research [19] . To investigate the role of astrocytes in neuronal development , a coculture system of astrocytes and neurons was adopted to mimic the situation of astrocytes and neurons in the brain . The neuronal dendrites were stained with the neuronal marker microtubule-associated protein 2 ( MAP2 ) . Based on our observations , the dendritic morphology was quite different in WT cortical neurons when cocultured with WT and KO astrocytes ( Figure 1C1 ) . We considered neurons with more than two short ( <50 µm ) dendrites as abnormal dendritic morphological neurons [22] . WT neurons cocultured with WT astrocytes exhibited only 7 . 6±1 . 9% cells with >2 short dendrites , whereas 72 . 2±6 . 1% cells with >2 short dendrites were observed when the WT neurons were cocultured with KO astrocytes ( Figure 1C2 ) . To evaluate more precisely , we performed a detailed morphological analysis , and found that the total dendritic length per cell was decreased by 74 . 2±5 . 8% when we compared neurons cocultured with KO astrocytes to those with WT astrocytes ( Figure 1C3 ) . A similar phenomenon was observed in KO neurons cocultured with WT and KO astrocytes ( Figure 1C4 ) . The cells with >2 short dendrites were 4 . 5±0 . 8% and 70 . 5±8 . 1% when cocultured with WT and KO astrocytes , respectively ( Figure 1C5 ) . The total dendritic length per cell of KO neurons was decreased by 85 . 3±6 . 4% when compared neurons cocultured with KO astrocytes to those with WT astrocytes ( Figure 1C6 ) . These results indicate that KO astrocytes alter the dendritic morphology of WT neurons , and WT astrocytes can prevent the abnormal morphology of KO neurons . Thus , astrocytes may play an important role in the development of FXS . To confirm that KO astrocytes resulted in abnormal dendritic morphology via the secretion of soluble factors , astrocytic conditioned medium ( ACM ) was generated to culture the neurons . The WT cortical neurons were plated in a low-density culture with WT and KO ACM respectively . Within 24 h , the neuronal survival was indistinguishable ( data not shown ) . At DIV 7 , we found a great variance of neuronal dendritic morphology between WT and KO ACM-treated neurons . The dendrites of the neurons treated with KO ACM were shorter and smaller; however , the neuronal densities were not different ( Figure 2A ) . Compared with the WT ACM-treated neurons , the number of cells with more than two short dendrites in the KO ACM-treated neurons increased by 58 . 1±19 . 8% ( Figure 2B ) , the total dendritic length per cell decreased by 62 . 9±9 . 5% ( Figure 2C ) . The 1∶1 ratio of WT and KO ACM showed a much better dendritic morphology and length than that in only KO ACM ( Figure 2B and 2C ) , suggesting that WT astrocytes can alleviate the neuronal abnormality caused by KO astrocytes . If the aberrant neuronal dendritic morphology is due to the lack of FMRP , the expression of FMRP in Fmr1 KO astrocytes is expected to rescue this deficit . The Fmr1 KO astrocytes expressed the enhanced green fluorescent protein ( EGFP ) when they were transfected successfully with FMRP constructs . As shown in the Figure 3A , the EGFP positive astrocytes were 52 . 3%±5 . 6% after 48 h of transfection . The transfection of FMRP vectors into KO astrocytes resulted in a 62 . 5%±7 . 2% FMRP protein expression of WT astrocytes ( Figure 3B ) . The ACM of FMRP-transfected KO astrocytes markedly altered the neuronal dendritic morphology as compared to the empty vector transfected KO ACM ( Figure 3C ) . The number of abnormal neurons was decreased by 42 . 1%±22 . 4% ( Figure 3D ) , and the total dendritic length per cell was increased by 139 . 7%±16 . 9% ( Figure 3E ) . In addition , the ACM of FMRP-transfected KO astrocytes also reversed the decreased levels of MAP2 , postsynaptic protein PSD95 , and glutamate receptor subunit GluR1 in the cultured neurons as compared to the empty vector transfected KO ACM-cultured neurons ( Figure 3F and 3G ) . Taken together , these results indicate that FMRP expression in astrocytes is important for neuronal development . Neurotrophic factors in the brain can promote neuronal survival and differentiation during development and participation in plasticity-related processes , such as NGF [23] , NT-3 [24] , and BDNF [25] . Thus , we detected some neurotrophic factors in the ACMs using ELISA kits . The ELISA kits used in the present study were designed for the total target proteins , not uniformly specific for mature and immature forms of the proteins . Interestingly , we found that BDNF , NGF , CNTF , and GDNF were almost unchanged in KO ACM compared with WT ACM . However , the NT-3 concentration in KO ACM ( 284 . 3±54 . 3 pg/ml ) was more than two folds of that in WT ACM ( 133 . 4±18 . 4 pg/ml , Figure 4A ) . At aged 3∼4 weeks old mice , the NT-3 levels in the cerebral cortex of Fmr1 KO mice were also higher ( 19 . 1±1 . 4 pg/mg protein ) than that of WT mice ( 11 . 1±1 . 2 pg/mg protein , Figure 4B ) . The NT-3 levels in older mice were not studied , because in adult tissues , neither WT nor KO astrocytes would express FMRP [19] , [26] . Unlike NT-3 , the other four neurotrophic factors remained unchanged both in vitro and in vivo ( Figure 4A and 4B ) . Additionally , the levels of NT-3 , BDNF , NGF , CNTF , and GDNF from the neuron culture medium were too low to detect . Therefore , excessive NT-3 released from the astrocytes of KO mice might be one of the causes for abnormal neuronal growth . To determine the reason for high levels of NT-3 released from Fmr1 KO astrocytes , the NT-3 mRNA levels of WT and KO astrocytes were examined . We found that the NT-3 mRNA levels in WT and KO astrocytes did not change ( Figure S1A ) . However , the NT-3 protein levels were higher in KO astrocytes than in WT astrocytes ( Figure S1B and S1C ) . These data suggest that the aberrant secretion of NT-3 in KO astrocytes is not caused by transcriptional regulation , but likely by protein synthesis . As we know , FMRP has multiple RNA-binding motifs and is involved in translational regulation [7] . To verify that NT-3 mRNA is a potential target of FMRP , we performed RIP to examine whether NT-3 mRNA directly binds with FMRP . We found the NT-3 mRNA in FMRP immunoprecipitates from WT astrocytes but not from Fmr1 KO astrocytes ( Figure 5C ) . A known FMRP-interacting mRNA , MAP1B [27]–[29] , was co-precipitated as the positive control ( Figure 5A ) , whereas a negative control mRNA ( GAPDH ) was not co-precipitated with the FMRP ( Figure 5B ) . The homology classes of BDNF and NGF mRNA did not bind with FMRP ( Figure 5D and 5E ) . Our findings indicate that NT-3 mRNA is part of the FMRP-mRNA complex and the translation of NT-3 mRNA is regulated by FMRP . To determine whether NT-3 is responsible for the dendritic disorder caused by KO ACM , we added exogenous NT-3 into WT ACM to observe the change in neuronal morphology . According to the difference between the NT-3 levels in WT and KO ACM , we set a low ( 150 pg/ml ) and a high ( 300 pg/ml ) dose of NT-3 in WT ACM . The neurons were stained with the dendritic protein MAP2 ( Figure 6A ) . We observed that the dendrites of the neurons treated with 150 pg/ml of NT-3 had a similar morphology to those of KO ACM-treated neurons , as shown by the smaller and shorter dendrites ( Figure 6B and 6C ) . Furthermore , the neurons treated with 300 pg/ml of NT-3 exhibited a worse developmental morphology ( Figure 6B and 6C ) . In contrast to NT-3 , the other two neurotrophins , NGF and BDNF , had no effects on neuronal development ( Figure S2A–S2C ) . The dendritic damage was validated further by the detection of the postsynaptic elements of excitatory synapses ( Figure 6D ) . The expressions of MAP2 , PSD95 , and GluR1 were significantly decreased in neurons treated with NT-3 as compared to the WT ACM-treated neurons . Moreover , this reduction was NT-3 dose-dependent ( Figure 6E ) . Because the major actions of NT-3 on neurons are produced by binding to the high affinity receptor tyrosine kinase C ( TrkC ) , we further studied the TrkC expression in neurons , and found that WT and KO neurons expressed similar levels of the TrkC ( Figure 6F ) . To provide direct evidence that excessive astrocyte-derived NT-3 inhibits dendritic development , we analyzed the effects of neutralizing antibody against NT-3 on neuronal dendritic development after KO ACM treatment . We found that the effect of NT-3 neutralization on neuronal dendritic development was dose dependent ( Figure 7A–7C ) . The classic bell-shaped curve depicted dendritic growth in response to NT-3 antibody ( Figure 7C ) . With the increase of dose , the neutralizing antibody to NT-3 gradually rescued dendritic growth , and 2 µg/ml of NT-3 antibody took the best effects to rescue the dendritic growth . However , high doses of NT-3 antibody ( 20 and 40 µg/ml ) exhibited less effect to the dendritic growth ( Figure 7B and 7C ) . This suggests that excessive NT-3 contributes to abnormal dendritic morphology while physiological levels of NT-3 are also necessary for dendritic development . To further evaluate the biological role of astrocyte-derived NT-3 in the dendritic disorder , we knocked down the NT-3 via shRNA infection . The KO astrocytes expressed the green fluorescent protein ( GFP ) when they were infected successfully with NT-3 shRNA . As shown in the Figure 8A , the GFP positive astrocytes were 92 . 5%±6 . 6% after 24 h of shRNA infection ( Figure 8A ) . This approach successfully reduced the NT-3 level by two shRNAs for NT-3 , while shRNA si-462 was incompetent to knockdown NT-3 expression ( Figure 8B ) . The NT-3 protein band intensity reduced to 31 . 6%±6 . 2% and 26 . 5%±3 . 6% of control by shRNAs si-508 and si-889 respectively ( Figure 8B ) . Consistently , the levels of the NT-3 decreased significantly to 186 . 0±9 . 5 pg/ml and 179 . 2±12 . 4 pg/ml in the ACM of KO astrocytes by shRNAs si-508 and si-889 respectively ( Figure 8C ) . As expected , the inhibition of KO astrocyte-derived NT-3 partially rescued the aberrant neuronal dendritic morphology ( Figure 8D ) . Compared to KO ACM treatment , the number of abnormal neurons was decreased by shRNAs si-508 and si-889 , but not by shRNA si-462 and the negative control ( Figure 8E ) . Consistently , the total dendritic length per cell was increased by shRNAs si-508 and si-889 , but not by shRNA si-462 and negative control ( Figure 8F ) . Next , the effect of NT-3 knockdown on the synaptic protein levels was observed . As shown in Figure S3 , the levels of MAP2 , PSD95 , and GluR1 were partially rescued in NT-3 shRNA-infected KO ACM-cultured neurons as compared to those in KO ACM-cultured neurons . These data indicate that interfering NT-3 expression in KO astrocytes can prevent abnormal neuronal growth in KO ACM . To determine whether the cognitive defects of Fmr1 KO mice may be related to the excessive NT-3 in the brain , we want to rescue the phenotype of Fmr1 KO mice by ACC local injection of negative shRNA infected WT astrocytes or NT-3 shRNA infected KO astrocytes ( Figure 9A ) . Ten days later , trace fear memory and the levels of NT-3 in the ACC were detected . First , mice were tested in the trace fear conditioning paradigm . Trace fear conditioned learning requires an intact ACC [5] , [30] . This paradigm differs from the classic delay paradigm in that the animal must sustain attention during the trace interval to learn the CS–US association [30] . The CS , an 80 dB white noise delivered for 15 s , was delivered 30 s before ( trace ) the US , a 0 . 7 mA scrambled foot shock . Mice were presented with 10 CS–trace–US trials with an intertrial interval ( ITI ) of 210 s . One day after training , mice received 10 CS–ITI trials in a novel chamber to test for trace fear memory . Before training , Fmr1 KO mice displayed similar baseline freezing compared with WT mice ( Figure 9B ) . WT mice successfully learned the trace fear conditioning after 10 CS–US pairings and showed increased freezing throughout the training session . Freezing was also increased in Fmr1 KO mice with ACC microinjection of negative shRNA infected WT astrocytes or NT-3 shRNA infected KO astrocytes ( Figure 9B ) . The average freezing during the intertrial intervals of the training session was significantly increased in Fmr1 KO mice with ACC microinjection of negative shRNA infected WT astrocytes or NT-3 shRNA infected KO astrocytes compared with Fmr1 KO sham mice ( Figure 9C ) . Similarly , Fmr1 KO mice with ACC microinjection of negative shRNA infected WT astrocytes or NT-3 shRNA infected KO astrocytes showed increased average freezing within the intertrial intervals of the testing session when compared with Fmr1 KO sham mice ( Figure 9C ) . Then , we found that levels of NT-3 were significantly decreased in the ACC of KO mice after local injection of negative shRNA infected WT astrocytes and the NT-3 shRNA infected KO astrocytes ( Figure 9D ) . These results suggest that down-regulation of NT-3 by microinjection of WT astrocytes or NT-3 shRNA infected KO astrocytes rescues the impairment at acquiring trace fear memory during training as well as in the expression of trace fear memory during testing on the following day .
FMRP is expressed in astrocytes , oligodendrocytes , and microglia in addition toneurons [19] , [31] . The role of astrocytes in the altered neurobiology of FXS has been first demonstrated by Jacobs and Doering , showing that Fmr1 KO astrocytes have profound effects on dendrites such as reduced length of dendrites and arbor area [20] . Using a coculture design , they found that hippocampal neurons exhibited abnormal dendritic morphology and a decreased number of presynaptic and postsynaptic protein aggregates when they were grown on astrocytes from a Fragile X mouse . Moreover , normal astrocytes could prevent the development of abnormal dendrite morphology and preclude the reduction of presynaptic and postsynaptic protein clusters in neurons from a Fragile X mouse . These experiments established a role for astrocytes in the altered neurobiology of FXS [20] . The further study shows that hippocampal neurons grown on Fragile X astrocytes exhibited delayed growth characteristics and abnormal morphological features in dendrites and synapses [32] . The present paper confirms these results and extends them to show that NT-3 is involved . In this study , astrocytic conditioned medium culture protocols were applied to demonstrate that KO astrocytes affect neuronal growth through altered secretion of soluble factors . Mutant products in astrocytes and microglia can damage neighboring neurons in some neurodegenerative disorders either by releasing toxic components or by mutant-mediated reduction in neuronal support functions [33] . NGF has been reported to decrease the survival of cultured cerebellar granule cells [34] or increase the number of cell deaths in the developing isthmo-optic nucleus [35] . Moreover , high levels of NT-3 are expressed in autism [36] and bipolar disorders [37] . In addition to neurons , astrocytes can represent an important local cellular source of neurotrophins , including NGF , BDNF , GDNF , CNTF , and NT-3 [38] , [39] . To investigate the causes for abnormalities of neuronal morphology , we hypothesized the absence or down-regulation of certain neurotrophic factors in KO ACM . The neurotrophic factors are a group of proteins that promote the survival and growth of neurons in the vertebral nervous system . However , in contrast with our expectation , we found higher levels of NT-3 in KO ACM and that excessive NT-3 was toxic to neuronal growth . It has been reported that exogenous NT-3 increases the total number of neuritis of neural plate explants but also the level of apoptosis in early neural development , and that blockade of NT-3 using an antibody reverses these effects [40] . In the present culture system , neuron-derived neurotrophic factors were undetectable . Thus , the significant finding of this study is the neurotoxicity observed only with excessive astrocyte-derived NT-3 from KO mice , which is traditionally considered a critical regulatory factor in cell proliferation , differentiation , and elimination of excess neurons produced during the course of normal brain development [41] . In addition , the overexpression of NT-3 was also detected in the cerebral cortex of Fmr1 KO mice and the levels of NT-3 in the brain tissue were lower than that in the culture medium . The diminution of the NT-3 difference in vivo and in vitro could be explained by the involvement of neurons and the state of the astrocytes . NT-3 is produced not only by astrocytes but also by neurons [38] , [39] . The levels of NT-3 in the brain can be influenced by hormones [42] and other members of the neurotrophin family [43] , [44] . The cultured astrocytes produce NT-3 [39]; however , in the brain ( in vivo ) , the reactive gliosis occurs in response to damage to the CNS and the content of neurotrophic factor in reactive gliosis is significantly higher relative to normal astrocytes [38] , [45] . Actually , the source of the NT-3 in the cerebral cortex samples in the present study cannot be distinguished . Two types of receptors that vary in terms of ligand binding specificity mediate the effect of neurotrophins . The low-affinity neurotrophin receptor p75 is capable of binding with all neurotrophins with equivalent affinity , whereas tyrosine kinase ( Trk ) family members exhibit ligand selectivity . The TrkC receptor is unique as it binds only with NT-3 and not with any other related ligands [46] . NT-3 at high concentration has also been reported to block the survival normally seen with BDNF [47] . In this study , we found that the neuronal damage by NT-3 was dose dependent , and the tremendous neurotoxicity was revealed by 300 pg/ml of NT-3 . The differences in the effects of trophic factors between neuronal populations may be due to the different properties inherent in the cells studied , i . e . varying expression levels of different Trk p75 receptors and different responses to environmental changes and lesions [35] . However , physiological levels of NT-3 are also necessary to dendritic development . By applying NT-3 neutralizing antibody , we found a classic bell-shaped curve of the dendritic growth in response to the dose of NT-3 antibody . With the increase of dose , NT-3 antibody gradually ameliorated dendritic growth . The peak dose was 2 µg/ml , higher doses of NT-3 antibody ( 20 and 40 µg/ml ) exerted less effects to the dendritic growth . This suggests that physiological levels of NT-3 are necessary; however , excessive levels are harmful to neuronal dendritic development . In this study , we used two biological methods to verify the function of NT-3 in neuronal development , i . e . the transfection of FMRP constructs into cultured Fmr1 KO astrocytes and knockdown of NT-3 using shRNAs . Both of them could partially rescue the neuronal dendritic morphology disorder . Further RNA-binding protein immunoprecipitation showed that NT-3 mRNA in FMRP immunoprecipitates in WT astrocytes . As we know , FMRP is associated with polyribosomes and involved in the translational efficiency and/or trafficking of certain mRNAs [7] . These results suggest that excessive NT-3 might be caused by a loss of translation repression of NT-3 mRNA due to the lack of the FMRP in Fmr1 KO astrocytes . In vitro neutralization against NT-3 provided direct evidence that excessive NT-3 contributes to abnormal dendritic morphology . However , the classic bell-shaped curve depicted dendritic growth in response to NT-3 antibody , suggesting that physiological levels of NT-3 are also necessary todendritic development . Further , in vivo ACC microinjection of WT astrocytes or NT-3 shRNA transfected KO astrocytes induced a significant reduction of the NT-3 level in the ACC and rescued the impairment of trace fear memory in the KO mice . Thus , the data presented here provide a possible explanation for the role of astrocytes in the abnormal neuronal dendritic development of FXS and may provide insights into the cellular mechanisms underlying Fragile X syndrome .
Fmr1 KO ( FVB . 129P2-Fmr1tm1Cgr/J; stock #4624 ) and control [FVB . 129P2-Pde6b+ Tyrc-ch/AntJ; stock #4828; hereafter referred to as wild type ( WT ) ] strain mice were obtained from The Jackson Laboratory . Mice were housed under a 12 h light/dark cycle with food and water provided ad libitum . All animal protocols used were approved by the Animal Care and Use Committee of the Fourth Military Medical University . The primary cultures of the cortical neurons were isolated from 15-day-old embryos ( E15 ) of Fmr1 WT and KO mice [14] . Twenty-four hours after plating with Dulbecco's Modified Eagle Medium ( DMEM ) containing 20% fetal bovine serum ( FBS ) ( Invitrogen , Carlsbad , USA ) , the medium was completely changed into a neurobasal medium with 2% B27 supplement ( Invitrogen ) . Half of the medium was changed every three days , and the neurons were used in the experiments at 7 days in vitro ( DIV ) . The primary astrocytes were isolated from either WT or Fmr1 KO one- ( P1 ) to two-day-old ( P2 ) pups using the differential adhesion method [48] . The dissociated cells were placed in a 25 cm2 tissue culture flask in DMEM containing 10% FBS and 2 mM of glutamine . When confluent ( after 6 to 7 days ) , the flasks were sealed and shaken at 250 rpm at 37°C for 16 h . Over 95% of the adherent cells were astrocytes as demonstrated by the anti-glial fibrillary acidic protein ( anti-GFAP ) immunostaining . The coculture of the neurons with astrocytes was performed as described previously [22] . The neurons were plated onto cover slips containing paraffin dots in neuronal plating medium . The cover slips were taken out after 4 h when the neurons adhered and then placed in 24-well culture plates containing a monolayer of astrocytes ( 50% to 60% confluence ) in the NB/B27 neuronal maintenance medium after inversion . The astrocytes were digested with 0 . 25% trypsin ( Invitrogen ) plus 0 . 02% ethylenediaminetetraacetic acid ( EDTA ) and then passaged to a 25 cm2 tissue culture flask . The cells were confluent after three days . The cultures were washed extensively with Hank's balanced salt solution ( HBSS ) ( Invitrogen ) , and the medium was replaced with NB/B27 culture medium to generate the ACM . After which , the ACM were collected at 7 DIV as previously described [22] . To culture the neurons in ACM , we isolated the cortical neurons and plated them on poly-L-lysine-coated plates containing the neuronal plating medium . The medium was kept for 4 h and then replaced by a 1∶1 mixture of ACM and neuronal maintenance medium to ensure that the neurons have adhered to the plate surfaces . Half of the medium was changed every three days throughout the cultures . The neurons were fixed in 4% paraformaldehyde for 20 min at room temperature after being cultured for seven days . After blocking for 30 min ( 5% BSA and 0 . 1% TritonX-100 ) , the cells were stained with anti-MAP2 ( 1∶1000 , Millipore , Billerica , USA ) overnight at 4°C followed by secondary Cy3-conjugated anti-rabbit antibody ( 1∶200 , Boster Bio-Technology , Wuhan , China ) . The astrocytes were similarly fixed and stained with anti-GFAP ( 1∶2000 , Millipore ) and secondary Alexa488-conjugated anti-mouse antibody ( 1∶800 , Invitrogen ) . The neurons and astrocytes were stained with Hoechst33258 ( 10 µg/ml , Beyotime Institute Biotechnology , Shanghai , China ) to observe their nuclei . The slides were observed using a confocal laser microscope ( FV1000 , Olympus , Tokyo ) , and the images were captured using Fluoview 1000 ( Olympus ) . For quantification of cell morphology , photomicrographs were taken randomly from each culture condition . The area of astrocytes and total dendritic length of isolated neurons were measured by ImageJ software [20] , [22] . The photography and analysis of immunoreactivity were performed in an investigator-blinded manner in three independent experiments . Each experiment used three or more cover slips to obtain the sample of isolated neurons for analysis . The cells were lysated in a RIPA lysis buffer ( Beyotime Institute Biotechnology ) and then briefly sonicated . Equivalent amounts of protein were resolved using 10% SDS-PAGE gel and transferred to the nitrocellulose membrane . After incubation with antibodies , the proteins were observed using enhanced chemiluminescence ( ECL , GE Healthcare Pharmacia , Uppsala , Sweden ) . The following primary antibodies were used: anti-FMRP ( 1∶1000; Millipore ) , anti-GFAP ( 1∶1000; Millipore ) , anti-MAP2 ( 1∶1000; Millipore ) , anti-PSD95 ( 1∶2000; Abcam , Cambridge , UK ) , anti-GluR1 ( 1∶300; Abcam ) , anti-NT-3 ( 1∶200; Santa Cruz , CA , USA ) , anti-TrkC ( 1∶200; Santa Cruz ) , and anti-β-actin ( 1∶10000; Sigma , St . Louis , MO ) . The secondary antibodies were horseradish peroxidase conjugated goat antibody to rabbit or mouse immunoglobulin ( 1∶10000; Boster Bio-Technology ) . The densitometric analysis of Western blots was conducted using a ChemiDoc XRS ( Bio-Rad , Hercules , CA ) and quantified using Quantity One version 4 . 1 . 0 ( Bio-Rad ) . The ACM samples were collected after brief centrifugation . The cerebral cortex tissues were normalized via Bicinchoninic Acid ( BCA ) Protein Assay to generate homogenates . The activation was evaluated by measuring the nerve growth factor ( NGF ) , neurorophin-3 ( NT-3 ) , brain-derived neurotrophic factor ( BDNF ) , glial cell-derived neurotrophic factor ( GDNF ) , and ciliary neurotrophic factor ( CNTF ) in the medium using enzyme-linked immunosorbent assay ( ELISA ) kits according to the manufacturer's instruction ( Westang , Shanghai , China ) . The concentration of these neurotrophic factors was detected using Denley Dragon Wellscan MK 3 and quantified using Ascent software for Multiskan . Full-length mouse Fmr1 was produced by reverse transcription-polymerase chain reaction ( RT-PCR ) using the following primers: 5′-CGGAATTC ( EcoRI ) GACGAGAAGATGGAGGAG-3′ and 5′-CCCTCGAG ( XhoI ) TACGGAAATGGTAGAGGA-3′ . The purified PCR product was cloned into the pIRES2-EGFP vector with the corresponding restriction enzyme sites ( EcoRI/XhoI ) . The recombinant DNA was confirmed by sequencing , and the expression of correctly sized proteins was confirmed via Western blot using an anti-FMRP antibody . The FMRP expression vector or an empty vector ( control ) were transfected using Lipofectamine LTX and PLUS Reagents ( Invitrogen ) according to the manufacturer's instructions . The WT and KO astrocytes were collected , and the total RNA was isolated from each sample using a Trizol reagent ( Invitrogen ) according to the manufacturer's instructions . The total RNA ( 2 µg ) was reverse transcribed using reverse transcriptase ( TaKaRa Biotechnology ) . The first strand cDNA was used as the template for real-time quantitative PCR analysis . The primers used were as follows: 5′-CATGTCGACGTCCCTGGAAATAG-3′ ( forward ) and 5′-GGATGCCACGGAGATAAGCAA-3′ ( reverse ) for NT-3 and 5′-TGTGTCCGTCGTGGATCTGA-3′ ( forward ) and 5′-TTGCTGTTGAAGTCGCAGGAG-3′ ( reverse ) for the internal quantitative control GAPDH . The mRNAs were detected using SYBR Green PCR Master Mix ( TaKaRa Biotechnology ) and an ABI PRISM 7500 Sequence Detection System ( Applied Biosystems , UK ) using the comparative threshold cycle method for relative quantification . The thermal cycling conditions were as follows: 95°C for 30 s , 40 cycles of 95°C for 5 s , and 60°C for 34 s . The RIP analysis was performed using the Magna RIP Kit ( Millipore Bioscience Research Reagents ) . The primary astrocytes were dispersed in an appropriate volume of complete RIP lysis buffer . The magnetic beads for immunoprecipitation were prepared and subsequently incubated for 30 min at room temperature with 5 µg anti-FMRP antibody ( Millipore Bioscience Research Reagents ) . The immunoprecipitation of the RIP lysate and beads-antibody complex was performed at 4°C overnight . After the protein degradation with proteinase K at 55°C for 30 min , the RNA was extracted using phenol/chloroform/isoamyl alcohol and precipitated with ethanol . RT-PCR then was performed using gDNA Eraser ( TaKaRa Biotechnology ) . The following primers were used: NT-3 , 5′-GGGGTGGGCGAGACTGAATG-3′ ( forward ) and 5′-TCCCTGGGTGCCTCTGCTTT-3′ ( reverse ) ; BDNF , 5′-TGCCAGTTGCTTTGTCTTCT-3′ ( forward ) and 5′-AGTGTCAGCCAGTGATGTCG-3′ ( reverse ) ; NGF , 5′-TGAAGCCCACTGGACTAAA-3′ ( forward ) and 5′- ACCTCCTTGCCCTTGATG-3′ ( reverse ) ; the positive control MAP1B , 5′-GGCAAGATGGGGTATAGAGA-3′ ( forward ) and 5′- CCCACCTGCTTTGGTATTTG-3′ ( reverse ) ; and the negative control GAPDH , 5′-TTAGCCCCCCTGGCCAAGG-3′ ( forward ) and 5′-CTTACTCCTTGGAGGCCATG-3′ ( reverse ) . The PCR products were separated and visualized on an agarose gel containing 5 g/l ethidium bromide . In an attempt to block the neurotoxicity of excessive astrocyte-derived NT-3 , neutralizing antibody to NT-3 ( anti- human NT-3 pAb , catalog No . G1651; Promega , Madison , WI , USA ) was added to the KO ACM cultured wells on the day of plating and for the entire time in cultures ( 7 days ) . Normal rabbit IgG ( Santa Cruz ) was administered as a negative control . The antibodies and normal rabbit IgG were resuspended in sterile PBS . The neutralizing specificity has been demonstrated previously [49] , [50] . In the present culture system , this neutralizing antibody was used at a final concentration of 0 . 02 , 0 . 2 , 2 , 20 , and 40 µg/ml , respectively . The NT-3 shRNAs lentiviruses were packaged by Genepharma ( Shanghai , China ) . Three sequences ( si-462: GGTCAGAGTTCCAGCCAATGA , si-508: GCAACAGAGACGCTA CAATTC , and si-889: GCAAACCTATGTCCGAGCACT ) were designed to reduce NT-3 expression compared with cells infected with lentivirus containing a nonsense control sequence ( negative: TTCTCCGAACGTGTCACGT ) . For the infection process , the astrocytes were plated onto 10 cm cell culture dishes and grown to 30% to 50% confluence before infection . The cells were incubated with lentivirus in serum-free conditions for 24 h at 37°C and then rinsed with PBS . After being incubated in DMEM containing 10% FBS for another 24 h , the cells were treated with NB/B27 and the ACM was collected . Before transplantation in vivo , the cultured WT and KO astrocytes were infected by negative and si-508 NT-3 lentivirus , respectively . Two days later , the infected astrocytes were carried with green fluorescence , then digested and resuspended at a concentration of 1×104 cells/µl in PBS . The experimental mice were 3∼4 weeks old , and divided into four groups , each group containing 5 mice . Two groups of Fmr1 KO mice were anesthetized with pentobarbital sodium ( 1 mg/20 g body weight , intraperitoneally ) and cells were stereotaxically microinjected into bilateral anterior cingulate cortex ( ACC; 1 mm anterior to the bregma , ±0 . 3 mm lateral from the midline , and 1 . 4 mm beneath the surface of the skull ) at a rate of 1 µl/min and at a volume of 1 µl/side , resulting in a dose of 1×104 cells per side . The other two groups were WT and KO mice sham operated controls , respectively . The mice were subjected to the same surgery but without cells in PBS . To confirm the localization and vitality of the injected cells in ACC , the brain was prepared from a surgical mouse and perfused with physiological saline , followed by fixative solution of 4% paraformaldehyde . The brain was post-fixed , sucrose-dehydrated and frozen-sliced at 20 µm thickness using a freezing microtome ( CM1950 , Leica , Germany ) . A coronal section including the injection's nick was selected and processed for green fluorescence observation . After ten days of surgery , trace fear conditioning was performed in an isolated shock chamber ( Med Associates , St . Albans , VT ) . The CS used was an 80 dB white noise , delivered for 15 s , and the US was a 0 . 7 mA scrambled footshock for 0 . 5 s . Mice were acclimated for 60 s and were presented with 10 CS–trace–US intertrial interval [31] trials ( trace , 30 s; ITI , 210 s ) . One day after training , mice were acclimated for 60 s followed by 10 CS–ITI trials ( ITI , 210 s ) in a novel chamber to test for trace fear memory [5] , [51] , [52] . All data were video recorded using FreezeFrame Video-Based Conditioned Fear System and analyzed by Actimetrics Software ( Coulbourn Instruments , Wilmette , IL ) . Average freezing for the baseline and for each ITI during the training and testing sessions was analyzed . Bouts of 1 . 0 s were used to define freezing . The data were expressed as mean±SEM . The statistical comparisons were performed via analysis of variance ( ANOVA ) . If the ANOVA was significant , post hoc comparisons were conducted using Tukey's test . In all cases , P<0 . 05 was considered statistically significant .
|
Fragile X syndrome is a form of inherited mental retardation in humans that results from expansion of a CGG repeat in the Fmr1 gene . Recent studies suggest that astrocytes play a role in neuronal growth . In this study , we find that astrocytes derived from a Fragile X model , the Fmr1 knockout ( KO ) mouse , inhibit the proper elaboration of dendritic processes of neurons in vitro . Excessive neurotrophin-3 ( NT-3 ) is released in the astrocytes from Fmr1 KO mice . Blockage of NT-3 by neutralizing antibodies and knockdown of NT-3 by using short hairpin RNAs ( shRNAs ) in Fmr1 KO astrocytes can rescue the neuronal dendritic development . In vivo experiments show that prefrontal cortex microinjection of WT astrocytes or NT-3 shRNA–infected KO astrocytes rescues the deficit of trace fear memory in KO mice . This study provides the evidence that a lack of FMRP leads to an overexpression of NT-3 , which reduces dendritic growth in neurons .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"neuroglial",
"development",
"neurobiology",
"of",
"disease",
"and",
"regeneration",
"developmental",
"neuroscience",
"cellular",
"neuroscience",
"neuronal",
"morphology",
"biology",
"neuroscience"
] |
2012
|
Excessive Astrocyte-Derived Neurotrophin-3 Contributes to the Abnormal Neuronal Dendritic Development in a Mouse Model of Fragile X Syndrome
|
Leptospirosis is a worldwide spread zoonotic and neglected infectious disease of human and veterinary concern that is caused by pathogenic Leptospira species . In severe infections , hemostatic impairments such as coagulation/fibrinolysis dysfunction are frequently observed . These complications often occur when the host response is controlled and/or modulated by the bacterial pathogen . In the present investigation , we aimed to analyze the modulation of the hemostatic and inflammatory host responses by the bacterial pathogen Leptospira . The effects of leptospires and their secreted products on stimulation of human intrinsic and extrinsic pathways of coagulation were investigated by means of altered clotting times , assembly and activation of contact system and induction of tissue factor . We show that both extrinsic and intrinsic coagulation cascades are modulated in response to Leptospira or leptospiral secreted proteins . We further find that the pro-inflammatory mediator bradykinin is released following contact activation at the bacterial surface and that pro-coagulant microvesicles are shed from monocytes in response to infection . Also , we show that human leptospirosis patients present higher levels of circulating pro-coagulant microvesicles than healthy individuals . Here we show that both pathways of the coagulation system are modulated by leptospires , possibly leading to altered hemostatic and inflammatory responses during the disease . Our results contribute to the understanding of the leptospirosis pathophysiological mechanisms and may open new routes for the discovery of novel treatments for the severe manifestations of the disease .
Leptospirosis is an infectious disease caused by pathogenic bacteria of the genus Leptospira [1 , 2] . In humans , infections are mainly acquired through contact with wild or domestic infected animals or exposure to contaminated soil or water [3 , 4] . It is estimated that more than 500 , 000 cases of leptospirosis occur annually world-wide [5] . Leptospires enter the host mainly via sodden or damaged skin or mucosa , followed by rapid dissemination through the blood stream . In the acute phase , or leptospiremia , bacteria may multiply in the circulation and spread into the surrounding tissue , being the kidneys and liver the preferential colonization sites . Soon after the host has mounted a specific immunological response , bacteria are cleared from blood , characterizing the immune or convalescent phase [2] . Infections can trigger a wide spectrum of clinical symptoms , varying from subclinical to severe manifestations . The most severe conditions known as Leptospirosis Pulmonary Hemorrhagic Syndrome and Weil’s disease , the last characterized by jaundice , hypotension , meningitis , kidney and multiple organ failure and hemorrhages , result in mortality rates up to 70% and 15% respectively [2 , 6] . The mechanisms of pathogenicity and virulence of the leptospires are still to be elucidated and the origin of pathophysiological leptospirosis symptoms and severity of disease remain virtually unknown [7–9] . During infection , inflammatory mediators from the microbe and/or host can induce complications by modulating the hemostatic equilibrium between the pro-coagulant and anticoagulant status of the host [10] . The coagulation cascade can be divided into two pathways , of which the extrinsic pathway is induced by tissue factor ( TF ) exposure and/or release and considered as the primary pathway coagulation [11] . The intrinsic pathway of coagulation , also referred to as the contact system , seems to play a secondary role in the processes . However , its activation can lead to a pro-inflammatory state via the release of bradykinin ( BK ) [12–14] . In severe bacterial infections , dysregulation of the host innate immune system and hemostasis can contribute to a fatal outcome . Notably , these complications often involve both pathways of the coagulation system [15] . We have previously reported that Leptospira are able to modulate the human fibrinolytic system . This interaction involves the capturing of human plasminogen on the surface of the microorganism , leading to increased pathogen-associated plasmin activity [16] . Leptospires also promote an imbalance of the normal fibrinolysis by enhancing the availability of plasminogen activators [17] . Bacterial membrane-associated plasmin can favor bacterial penetration , and also enhance immune evasion activity , a critical condition during hematogenous dissemination [17–19] . Some studies report ongoing fibrinolysis , activation of coagulation , impaired anticoagulation and thrombocytopenia during leptospirosis , while the involvement of disseminated intravascular coagulation ( DIC ) is controversial [20–24] . As leptospirosis can lead to serious hemorrhagic syndromes and to date no mechanism responsible for this is known , it is generally believed that impaired hemostasis through coagulation/fibrinolysis dysfunction might be involved . Thus , the present investigation aims to further decipher the role of these systems in leptospirosis .
Confirmed leptospirosis human serum samples were obtained from Instituto Adolfo Lutz sera collection , Sao Paulo , Brazil , and were donated for research purposes . We had no access to any patient data . The use of serum samples from Adolfo Lutz sera collection was waived from official ethics approval by the Ethical Committee for Human Research of Universidade de Sao Paulo , which rules that this work does not involve procedures regulated by CONEP/Brazil n° 466/2012 . We used paired samples from the same patients at the early phase ( MAT negative , negative microscopic agglutination test ) and convalescent phase of leptospirosis ( MAT positive ) . The MAT was performed according to Faine et al . [3] . In brief , an array of serovars of Leptospira spp . as antigens were employed . A laboratory-confirmed case of leptospirosis was defined by a four-fold microagglutination titer rise between paired serum samples . MAT was considered negative when the titer was below 100 . Virulent low-passage L . interrogans serovar Copenhageni strain L1-130 , non-virulent culture-attenuated L . interrogans serovar Copenhageni , and saprophytic L . biflexa serovar Patoc were kindly provided by Dr . Mathieu Picardeau ( Institute Pasteur , France ) . Leptospires were cultured at 28°C in Elinghausen-McCullough-Johnson-Harris ( EMJH ) medium ( BD , Difco ) supplemented with 10% Leptospira enrichment EMJH ( BD , Difco ) , 0 . 3 g/L peptone ( BD , Difco ) and 0 . 2 g/L meat extract ( Sigma-Aldrich ) . Blood samples were drawn from healthy volunteers in Vacutainer tubes ( Becton Dickinson ) containing 1/9 volume 109 mM sodium citrate , pH 7 . 4 . Blood was used immediately or centrifuged for plasma separation . Informed consent was obtained from all healthy blood donors . Mid-log phase culture Leptospires were enumerated , washed twice with HEPES buffer ( 115 mM NaCl , 1 . 2 mM CaCl2 , 1 . 2 mM MgCl2 , 2 . 4 mM K2HPO4 , 20 mM HEPES ) and resuspended in HEPES buffer containing 50 μM ZnCl2 ( HEPES-ZnCl2 ) . 300 μL of the cells suspensions ( 5x109 , 1x109 , 1x108 or 1x107 cells/mL ) or mid-log phase culture supernatants were added to the same volume of blood or plasma . Fresh culture medium or buffer alone were employed as controls . After incubation for 0 . 5 , 1 , 2 or 4 h at 37°C , 50 μL 30 mM CaCl2 were added to 100 μL of the samples , and the recalcification clotting times ( i . e . after calcium addition ) were measured in a semi-automatic ball coagulometer ( MC10plus , Amelung ) . Alternatively , as a strategy to analyze the plasma after bacterial interaction with coagulation factors , bacteria were removed by centrifugation ( 4000 x g , 15 min ) after the incubation in plasma and the recalcification clotting times of the supernatants were measured . The remaining bacterial pellets were resuspended in 300 μL HEPES-ZnCl2 , mixed with fresh plasma , and the recalcification clotting times immediately measured . To measure aPTT ( activated partial thrombin time ) , PT ( prothrombin time ) , and TCT ( thrombin clotting time ) , 1x109 bacteria/mL were washed and resuspended in buffer . The samples were mixed with the same volume of human plasma , and after 30 min incubation at 37°C , bacteria were removed by centrifugation . To trigger the clotting reactions , 100 μL of the resulting supernatants were mixed with 20 μL PT ( TriniCLOT PT , Trinity Biotech ) or 50 μL TCT reagents ( 3 . 4 IU/mL thrombin ) , and the times for plasma to clot were recorded . Alternatively , 100 μL of the samples were mixed with 50 μL aPTT reagent ( DAPTTIN TC—Technoclone ) and after 200 s incubation , 50 μL 30 mM CaCl2 were added to initiate the clotting reaction . Human PBMCs were purified from citrated blood using Lymphoprep gradient ( AXIS-SHIELD ) , according to the manufacturer instructions . Equal volumes of PBMCs suspensions ( 5x106 cells/mL ) and leptospires ( 2x107/mL ) or bacteria culture supernatants were mixed and incubated at 37°C for 2 h . Buffer alone or EMJH were used as controls . The PBMCs were pelleted by centrifugation and resuspended in 600 μL normal plasma or FVII-deficient plasma diluted 50% in HEPES-ZnCl2 . Normal plasma was used to verify the general status of the coagulation , while FVII-deficient plasma as a control to exclude the activity of the TF/extrinsic pathway of coagulation . The recalcification clotting times were measured by addition of 50 μL 30 mM CaCl2 to 100 μL of sample . Alternatively , the aPTTs were measured in order to verify the status of the intrinsic ( contact system ) and common pathways of coagulation . aPTTs were obtained by addition of 50 μL aPTT reagent to 50 μL of sample followed by 200 s incubation and addition of 50 μL 30 mM CaCl2 . To evaluate the assembly and activation of the contact system at the bacterial surface , we used the chromogenic substrate S-2302 , specific for plasma kallikrein ( PK ) . To further assess the activity of downstream activation of the final common coagulation pathway , we used the chromogenic substrates specific for factor Xa ( FXa ) and thrombin ( S-2765 and S-2238 , respectively ) . All substrates were purchased from Chromogenix . 3x109 bacteria were washed and resuspended in 300 μL HEPES-ZnCl2 . Bacterial suspensions were added to equal volumes of plasma with or without 100 μg/mL of the peptide H-D-Pro-Phe-Arg-chloromethylketone , a specific inhibitor of the contact system by interfering with factor XII ( FXII ) and PK activities [25] , ( N-1210 , Bachem ) ( or buffer , as control ) . After incubation at 37°C for 30 min , bacteria were washed three times and resuspended in 450 μL buffer . Chromogenic substrates ( 50 μL of 4 mM solution ) were added to 150 μL aliquots of bacterial suspensions . After 30 min incubation at 37°C , cells were removed and the optical densities ( 405 nm ) of the supernatants were measured . Leptospira suspensions ( 3x108 in 300 μL HEPES-ZnCl2 ) were mixed with equal volumes of human plasma with or without 100 μg/mL N-1210 . After 30 min incubation at 37°C , bacteria were pelleted and extensively washed . To elute the bacteria-bound proteins , 100 μL 0 . 1 M glycine ( pH 2 . 0 ) were added to the bacterial pellets . Bacteria were removed by centrifugation and after addition of 10 μL 1 M Tris pH 7 . 4 , the supernatants were subjected to 10% SDS-PAGE . The gels were transferred to membranes and probed with antibodies against ( FXII ) or high molecular weight kininogen ( HK ) . After incubation with HRP-conjugated antibodies , reactivity was detected by chemiluminescence . Bacterial suspensions ( 1x109 cells/mL in HEPES-ZnCl2 ) were incubated with equal volume of normal or kallikrein-deficient plasma for 30 min at 37°C . Some samples received the addition of N-1210 ( 100 or 25 μg/mL ) or the cysteine proteinase inhibitor E-64 ( 100 μg/mL ) . Bacteria were washed , resuspended in buffer , incubated for 30 min at 37°C and centrifuged . The supernatants were collected for BK release determination by an ELISA kit ( ab136936 , Abcam ) . Quantification of TF and TFPI from human sera samples were performed using commercial ELISA kits ( ab108903 and ab108904 , Abcam ) , according to the manufacturer instructions . For TF and TFPI determinations , we used paired serum samples from 10 and 12 patients , respectively . Pooled leptospirosis patients sera or normal human sera ( 200 μL ) were centrifuged ( 20 , 800 x g , 30 min ) . Supernatants were disposed and the MVs were washed two times in 900 μL PBS . Following centrifugation ( 20 , 800 x g , 30 min ) , 800 μL of supernatants were removed , and the MVs were resuspended in the remaining buffer . For MVs isolation , we used paired sera from 10 patients , pooling 5 MAT negative and 5 MAT positive for two independent isolations . For the isolation of PBMCs-derived MVs , the cells were stimulated with leptospires or culture supernatants for 24 h at 37°C . After centrifugation to remove the cells and bacteria , the supernatants were subjected to MVs isolation as described above . TF was labeled with colloidal gold as described earlier [26] . MVs preparations were mixed with gold-labeled TF and processed for negative staining [27] .
To analyze whether leptospires can influence the human coagulation cascade , bacteria were incubated with citrated human blood and the recalcification clotting times were determined . The two pathogenic strains ( virulent and culture attenuated L . interrogans ) and the non-pathogenic strain ( L . biflexa ) reduced to the same extent the clotting times in a dose and time-dependent manner ( Fig 1A and 1B ) . Notably , incubation of bacteria with human plasma instead of human blood did not alter the recalcification clotting times ( Fig 1C ) . Together these data suggest that clotting is mainly triggered by activation of the extrinsic pathway of coagulation , as it is dependent on the cellular components of human blood . In the next experiments bacteria were incubated with human plasma for 30 min and then removed by centrifugation . After calcium addition to the remaining plasma a dramatic and strain-independent increase of the recalcification clotting times was noted at a Leptospira load of 1x109 bacteria/mL and when concentrations were 5x109 bacteria/mL clot formation was completely prevented ( Fig 1D ) . These findings suggest that a depletion of coagulation factors possibly via binding to the bacteria surface or degradation/consumption is responsible for the impairment of normal clotting . To better understand the coagulation disturbances caused by Leptospira , bacteria ( 1x109/mL ) were mixed with human plasma , removed by centrifugation and clotting times measuring activation of the extrinsic , intrinsic , and common pathway of coagulation were determined in the remaining supernatants . Prolonged aPTT and PT , but not TCT , values were observed for all three strains as compared to plasma samples incubated with buffer alone ( Fig 2A–2C ) . We additionally tested whether Leptospira strains influence the aPTTs , PTs , or TCTs , showing that also under these conditions normal clotting was impaired ( Fig 2D ) . Together our data suggest that coagulation is induced by an up-regulation of TF at cellular level , while the interaction of the bacteria with coagulation factors decelerates the time to clot formation . When bacterial culture supernatants were used to modulate clotting in human blood and plasma , results similar to the ones obtained with the bacteria were observed . Supernatants from the three strains reduced the recalcification clotting times of blood and plasma ( S1 Fig ) . Thus , leptospires have both secreted and surface-associated products which are able to modulate the coagulant state in human blood and plasma . Having shown that Leptospira and culture supernatants can trigger pro-coagulant activities in human blood , we next investigated whether this effect is caused by an up-regulation of TF . Bacteria and culture supernatants were incubated with isolated human PBMCs . Treated and untreated cells were added to fresh human plasma and the recalcification clotting times were determined . All three strains and their respective supernatants rendered PBMCs into a pro-coagulant state . The data further reveal that clotting is controlled by the extrinsic pathway of coagulation , because only decreased times were measured when normal but not FVII-deficient plasma was used ( Fig 3A and 3B ) . When stimulated or non-stimulated cells were added to normal or FVII-deficient plasma and aPTTs measured to verify the status of the intrinsic pathway of coagulation , no differences were noted ( Fig 3C and 3D ) , confirming that the pro-coagulant activity is mainly caused by the up-regulation of TF on PBMCs . Our results showing that addition of Leptospira to human plasma impairs the intrinsic pathway of coagulation , implies an interaction of the bacteria with factors of the contact system . To address this issue , bacteria were incubated with human plasma , washed , and the binding and activation of contact factors at the bacterial surface was assessed by specific chromogenic substrates . The assembly and activation of the contact system occurs at the surface of the three strains tested and resulted in not only a hydrolysis of the substrates specific for plasma kallikrein/FXII ( Fig 4A ) , assessing the status of the intrinsic pathway of coagulation , but also of substrates specific for FXa ( Fig 4B ) and thrombin ( common pathway ) ( Fig 4C ) . Interestingly , activation seems to correlate with the virulence of the strains tested , as L . interrogans L1-130 was the most potent followed by the non-virulent culture-attenuated L . interrogans and the non-pathogenic L . biflexa strains . When the experiments were conducted in the presence of a specific inhibitor of FXII and PK , peptide N-1210 [25] , a strong decrease of the PKa activity was measured and also the activity of FXa and thrombin was significantly reduced . Together the results show that leptospires are able to assemble and activate the contact system at their surface , leading to an activation of the up-stream coagulation factors i . e . FXa and thrombin . The data further suggest that the extent of activation may correlate with the virulence of these strains . To further analyze the binding and processing of contact proteins at the leptospiral surface , bacteria were incubated in human plasma in the absence or presence of N-1210 . Bound proteins were eluted and subjected to Western blot analysis using antibodies against FXII and HK . Fig 5A depicts FXII is recovered in its active form from the bacterial surface of all strains when plasma was incubated in the absence of peptide N-1210 , while in the presence of the inhibitor , activation of FXII was notably prevented . When analyzing the degradation pattern of HK under the same experimental conditions , similar findings were obtained , implying that also HK is processed at the bacterial surface of the three strains ( Fig 5B ) . To test whether HK cleavage is followed by the release of BK , the virulent L . interrogans strain and human plasma were incubated for 30 min and the release of BK from the bacterial surface after was measured after a washing step . Incubation of bacteria with normal plasma resulted in a massive BK release , which was not seen when PK-deficient plasma was used or plasma was replaced by buffer ( Fig 5C ) . Addition of N-1210 during the plasma incubation step resulted in a dose-dependent inhibition of BK release , while co-incubation with a cysteine proteinase inhibitor ( E64 ) had no significant effect . Taken together , these results indicate that the contact system is activated at the bacterial surface leading to a subsequent activation of further coagulation factors and the generation of BK . Next , we wished to analyze the coagulative state of plasma samples from Leptospira patients . We measured TF concentrations in samples from initial phase ( MAT negative ) and convalescent phase ( MAT positive ) from the same patients . We found the highest medium TF concentrations in MAT negative ( 134 . 79±54 . 98 pg/mL ) and MAT positive ( 97 . 33±30 . 66 pg/mL ) patients compared to the levels in sera from or healthy individuals ( below detectable values ) ( Fig 6A ) . Although there is no significant statistical difference between the two groups , we also observed that there is a general trend showing that TF levels tend to decline during disease progression ( Fig 6B ) . As TF activity is regulated by tissue factor pathway inhibitor ( TFPI ) , we determined the concentration of TFPI in these samples . There were no significant differences of TFPI levels in the two patients groups ( MAT negative 31 . 78±5 . 48 ng/mL; MAT positive 29 . 66±7 . 84 ng/mL ) compared to control ( 28 . 34±1 . 97 ng/mL ) ( Fig 6C ) . These results further indicate the pro-coagulant state during leptospirosis . MVs are membrane vesicles that are released by many cell types and have been recently implicated in a number of biological processes , such as hemostasis , inflammation and host defense , contributing to the pathogenesis of many diseases [28 , 29] . We therefore decided to determine the coagulative state of MVs released from purified human PBMCs stimulated with leptospires or culture supernatants . The PBMCs-derived MVs were added to human plasma , and the recalcification clotting times were determined . Both bacterial cells and secreted products were able to stimulate the release of pro-coagulant MVs by monocytes , as observed by the decreased recalcification clotting times when compared to the controls ( PBMCs stimulated with buffer or EMJH alone ) ( Fig 7A ) . We also analyzed the coagulative state from MVs purified from leptospirosis patients’ sera . MVs were added to plasma , and the recalcification clotting times with or without PT reagent were determined . Although no significant changes were observed for the recalcification time ( S2 Fig ) , a decrease in the clotting times was observed for both MAT negative and MAT positive samples in the presence of PT reagent ( Fig 7B ) . By immunoelectron microscopy , we observed that MVs derived from PBMCs stimulated with leptospires or its culture supernatants are smaller in size and present higher exposure of TF than the controls stimulated with PBS or EMJH ( Fig 8A–8H ) . A similar pattern was seen for patients-derived MVs , where a higher number of small nanovesicles harboring more TF than normal control were found for both early and convalescent-phases of the disease ( Fig 8I–8N ) . Together , our results indicate higher activity of the extrinsic pathway of coagulation , suggesting that during leptospirosis , pro-coagulant MVs harboring TF are shed from activated monocytes .
As soon as leptospires enter the host , mainly through skin or mucosa , the pathogen starts to disseminate throughout the body . Given the extremely high number of bacteria in the circulation [30] , the host immune system can ultimately trigger exaggerated systemic responses leading to severe complications . Such complications can be caused by a massive release of cytokines or dysfunction of the coagulation system [31] . In cases of excessive inflammation , leptospirosis patients may then suffer from high fever and tubulo-interstitial nephritis with interstitial edema and containing massive immune cells infiltrates [32 , 33] , whereas an impairment of the coagulation system is often associated with severe bleeding complications . Coagulation is a tightly controlled chain of events that can be initiated via an extrinsic or intrinsic pathway . While the extrinsic pathway is essential for normal clot formation , the intrinsic , also known as the contact system , plays a secondary role in hemostasis . However , the latter pathway is involved in triggering inflammatory reactions through the formation of the pro-inflammatory peptide BK [14 , 34–36] . Miragliotta and colleagues [37] have previously shown that monocytes incubated with Leptospira were able to shorten the recalcification time of normal plasma , but were ineffective in factor VII deficient plasma , suggesting the generation of TF-like activity . Accordingly , our results show an induction of TF-mediated pro-coagulant activity by Leptospira and their secreted products under in vitro conditions and increased TF levels in the sera of patients suffering from leptospirosis . These findings point to an important role of TF in the induction of the pro-coagulant state during leptospirosis . Moreover , our data further demonstrate that MVs isolated from patients samples have pro-coagulant properties , exhibiting different size characteristics and harboring more TF than healthy controls . Both results support the notion that systemic activation of coagulation can take place during leptospirosis and may eventually lead to severe coagulation disturbaces including DIC , though DIC occurence is controversial in leptospirosis [23 , 24 , 38] . Within the last two decades there are many studies reporting an involvement of the contact system in the pathophysiology of sepsis [39] . It has been shown that the assembly and activation of the contact system at the surface of human pathogens , such as Streptococcus pyogenes , Escherichia coli , and Salmonella spp . , can induce a massive release of BK , that in turn may lead to increased vascular permeability and plasma leakage and is followed by bacterial dissemination and massive inflammatory reactions [39–41] . These findings are in line with our results showing that leptospires provide a surface that allows the recruitment and activation of the contact system . The release of BK during leptospirosis may therefore constitute an important mechanism that considerably contributes to inflammatory reactions during the disease . Under systemic conditions this may contribute to some of the leptospirosis symptoms like fever , pain and hypotension . In addition , inflammation of blood vessels and vasculitis are conditions often associated to leptospirosis [42] , and thus our results provide a plausible explanation for the molecular mechanisms behind these complications . Apart from triggering inflammatory reactions , kinins can induce the release of tissue-type plasminogen activator ( tPA ) from endothelial cells [43 , 44] . Our previous studies have shown that elevated tPA levels can be measured in human leptospirosis , especially in the early phase of the disease [17] . As BK is an early inflammatory mediator , these findings support the idea that BK release could be involved in the mobilization of tPA during the hematogenous leptospirosis phase . Thus our results can provide an explanation for the hemostatic imbalance of leptospirosis by fibrinolytic activation . Bleeding disorders are common complications in severe leptospirosis and there are many evidences emphasizing that activation of coagulation , impaired anticoagulation , and activation of fibrinolysis play a significant role in these processes [17 , 20–24] . Although incipient , studies revealing mechanisms by which leptospires interfere with host immune responses and hemostasis have enlightened our understanding of Leptospira host-pathogen interactions in the last few years [19 , 45] . Our results further contribute to an increased knowledge of the pathophysiological mechanisms in leptospirosis and uncover new mechanisms by which this pathogen interferes with the host response and gives rise to pathological signs . Here we show that both pathways of the coagulation system are modulated by leptospires and this may open new directions for diagnosis and treatment in the severe manifestations of leptospirosis .
|
Leptospirosis is one of the most relevant and spread zoonotic and neglected infectious diseases affecting humans and other mammals , and is caused by pathogenic bacteria of the genus Leptospira . During infectious diseases , when bacterial pathogens control and/or modulate the host response , impaired hemostasis and inflammation are frequently observed . Here we studied the effects of leptospires and their secreted products on stimulation of human intrinsic and extrinsic pathways of coagulation , showing that both coagulation cascades are modulated in response to Leptospira or leptospiral secreted proteins . We further find that activation of the coagulation cascades culminates in the release of the pro-inflammatory mediator bradykinin and noted an induction of pro-coagulant microvesicles . These findings contribute to a better understanding of the local and systemic hemostastic complications during leptospirosis . Collectively , our results show how leptospires can affect host responses , possibly leading to altered host responses during the disease and giving rise to the leptospirosis symptomatology .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"inflammatory",
"diseases",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"leptospira",
"pathology",
"and",
"laboratory",
"medicine",
"coagulation",
"factors",
"pathogens",
"tropical",
"diseases",
"microbiology",
"bacterial",
"diseases",
"neglected",
"tropical",
"diseases",
"bacteria",
"bacterial",
"pathogens",
"infectious",
"diseases",
"zoonoses",
"proteins",
"medical",
"microbiology",
"microbial",
"pathogens",
"blood",
"plasma",
"leptospirosis",
"hematology",
"biochemistry",
"blood",
"coagulation",
"blood",
"anatomy",
"physiology",
"biology",
"and",
"life",
"sciences",
"leptospira",
"interrogans",
"organisms"
] |
2016
|
Modulation of Hemostatic and Inflammatory Responses by Leptospira Spp.
|
As a master regulator of jasmonic acid ( JA ) –signaled plant immune responses , the basic helix-loop-helix ( bHLH ) Leu zipper transcription factor MYC2 differentially regulates different subsets of JA–responsive genes through distinct mechanisms . However , how MYC2 itself is regulated at the protein level remains unknown . Here , we show that proteolysis of MYC2 plays a positive role in regulating the transcription of its target genes . We discovered a 12-amino-acid element in the transcription activation domain ( TAD ) of MYC2 that is required for both the proteolysis and the transcriptional activity of MYC2 . Interestingly , MYC2 phosphorylation at residue Thr328 , which facilitates its turnover , is also required for the MYC2 function to regulate gene transcription . Together , these results reveal that phosphorylation-coupled turnover of MYC2 stimulates its transcription activity . Our results exemplify that , as with animals , plants employ an “activation by destruction” mechanism to fine-tune their transcriptome to adapt to their ever-changing environment .
Plants are continuously challenged by various biotic and abiotic stresses with diverse modes of attack . In response to an attack , plant cells undergo dramatic transcriptional reprogramming to efficiently coordinate the activation of attacker-specific immune responses so that the optimal resistance is attained . Equally importantly , when the attacking alarm is relieved , plants cells must effectively suppress their immune responses at the right time to minimize the cost of defense . Therefore , plant cells have involved elaborate regulatory mechanisms to keep defense-related gene transcription tightly in check . Among the best-characterized molecular signals regulating plant immune responses is the jasmonic acid ( JA ) family of oxylipins , which orchestrate genome-wide transcriptional reprogramming of plant cells to coordinate defense-related processes . Much of our understanding of the JA signal transduction pathway has come from the recent elucidation of the molecular details of JA-regulated gene transcription through MYC2 , a basic helix-loop-helix ( bHLH ) -type transcription factor that regulates diverse aspects of JA responses [1]–[4] . At low JA levels , the transcriptional activity of MYC2 is repressed by JASMONATE ZIM DOMAIN ( JAZ ) proteins , which recruit TOPLESS ( TPL ) to form a transcriptional repressor complex through the adaptor protein NOVEL INTERACTOR OF JAZ ( NINJA ) [5]–[7] . A battery of stresses , including mechanical wounding , insect attack and pathogen infection , triggers a rapid increase of cellular JA levels . Synthesized JA is conjugated with isoleucine to form the active hormone JA-Ile , which is perceived by its receptor CORONATINE INSENSITIVE1 ( COI1 ) , an F-box protein that forms an E3 ubiquitin ligase [6] , [8]–[11] . JA-Ile acts as a “molecular glue” to stimulate the interaction between COI1 and JAZs , which bring JAZs for degradation and therefore relieves their repression effect on MYC2 [5] , [6] . Two MYC2-like bHLH-type transcription factors , MYC3 and MYC4 , were also able to interact with JAZs and act additively with MYC2 in the regulation of JA-signaled immune responses [12] . Although MYC2-mediated transcriptional regulation plays a central role in different aspects of JA-mediated immunity , how MYC2 itself is regulated at the protein level remains elusive . In mammalian , phosphorylation and the ubiquitin-proteasome system ( UPS ) -mediated proteolysis are prominent posttranslational mechanisms that control transcription factors [13] . One of the most extensively studied transcription factors whose activity is under the control of UPS-mediated proteolysis is the Myc oncoprotein , which is a bHLH-type transcription factor that shares structural features with MYC2 [14]–[16] . Myc is highly unstable and , surprisingly , the Myc degron–region that signals Myc destruction , is closely overlapped with its transcriptional activation domain ( TAD ) [17] , [18] , revealing a functional connection between protein destruction and gene activation . A growing body of evidence demonstrates that this functional overlap between degrons and TADs is not unique to Myc , but reflects a general phenomenon for most unstable transcription factors , leading to the “activation by destruction” hypothesis , in which the UPS-mediated turnover of transcription factors is essential for their ability to activate gene transcription [13] , [19] , [20] . Whereas “activation by destruction” is a general phenomenon in the mammalian system , evidence that this paradigm is also involved in the regulation of transcription factors in the plant kingdom is lacking . In this report , we investigated if MYC2 is regulated by posttranscriptional mechanisms . Our findings revealed that UPS-mediated proteolysis is involved directly and mechanistically in the regulation of MYC2 and demonstrate that plants employ proteolysis-coupled transcription as a mechanism to control their responses to various environmental stresses .
To investigate the mechanism by which MYC2 differentially regulates distinct subsets of JA responses , we followed the time-course of MeJA-induced expression of wound- and pathogen-responsive genes in wild type ( WT ) and the myc2-2 mutant [2] . For this analysis , we select LIPOXYGENASE2 ( LOX2 ) as a representative marker gene for wound response [21] and the plant defensin gene PDF1 . 2 as a representative marker gene for pathogen response [22] . Quantitative real-time PCR ( qRT-PCR ) assays indicated that , in MeJA-treated WT seedlings , LOX2 mRNA levels showed a pronounced increase at 3 h and reached to a maximum at 6 h; LOX2 mRNA levels then showed a tendency of reduction and returned to basal levels at 48 h after MeJA treatment ( Figure 1A ) . Parallel experiments indicated that , in MeJA-treated WT seedlings , PDF1 . 2 mRNA remained at basal levels until 12 h and reached to a maximum at 48 h after treatment ( Figure 1B ) . These results demonstrate that , JA-mediated induction of wound-responsive genes , which are positively regulated by MYC2 , occurs relatively early . In contrast , JA-mediated induction of pathogen-responsive genes , which are negatively regulated by MYC2 , occurs relatively late . We then examined JA-induced accumulation of MYC2 at both mRNA and protein levels . As shown in Figure 1C , MYC2 mRNA levels quickly reached to a maximum at 0 . 5 h after MeJA treatment then showed a tendency of reduction in the duration of the experiment . We then used the MYC2-4myc-15 plants , which express a functional MYC2-4myc fusion protein ( Figure S1 ) , to examine the MeJA-induced accumulation kinetics of the MYC2 fusion protein . As shown in Figure 1D , upon MeJA treatment , the MYC2-myc fusion protein showed an obvious induction at 0 . 5 h , maintained at very high level from 3 h to 12 h , and exhibited a tendency of reduction from 24 h to 48 h ( Figure 1D ) . Therefore , high accumulation of the MYC2 protein correlates with peaked expression of early wound-responsive genes , whereas low accumulation of the MYC2 protein correlates with peaked expression of late pathogen-responsive genes . These results implicate that temporal regulation of MYC2 protein accumulation is important for its function . Previous studies revealed two mechanisms by which MYC2 activates the expression of early wound-responsive genes including LOX2 , TAT1 and VSP1 . First , MYC2 activates LOX2 and TAT1 transcription by directly binding to their promoters [23]; Second , MYC2 directly activates the expression of intermediate transcription factors such as ANAC019 and ANAC055 [24] , which , in turn , activate the expression of VSP1 [25] . It was reported that the effect of MYC2 on the transcription of pathogen-responsive genes is mainly achieved through direct regulation of a spectrum of intermediate transcription factors [3] . For example , it was shown that the negative regulation of PDF1 . 2 expression by MYC2 is achieved through directly suppression of EHYLENE RESPONSE FACTOR 1 ( ERF1 ) [3] . We provided several lines of evidence supporting that , in addition to ERF1 , the transcription factor OCTADECANOID-RESPONSIVE ARABIDOPSIS AP2/ERF-domain protein 59 ( ORA59 ) , which directly binds the promoter of PDF1 . 2 [26] , is also involved in MYC2-mediated suppression of PDF1 . 2 expression . First , MeJA-induced expression levels of ORA59 was dramatically increased in myc2-2 than those in WT ( Figure S2 ) , indicating that MYC2 negatively regulates MeJA-induced expression of ORA59 . Second , chromatin immunoprecipitation ( ChIP ) assays using the previously described 35Spro:MYC2-GFP plants [27] indicated that MYC2 associates with a G-box hexamer ‘CACGTG’ in the ORA59 promoter ( Figure 2A and 2B ) . Third , DNA electrophoretic mobility shift assays ( EMSA ) indicated that a MYC2-maltose binding protein ( MBP ) fusion protein binds the ORA59 promoter sequence in a G-box-dependent manner ( Figure 2A and 2C ) . Finally , using the transient expression assay of Nicotiana benthamiana leaves , we verified the repression effect of MYC2 on the expression of a reporter containing the ORA59 promoter fused with the firefly luciferase gene ( LUC ) ( Figure 2D–2F ) . Together , these results indicate that ORA59 is a member of the intermediate transcription factors involved in MYC2-mediated suppression of PDF1 . 2 expression . The existence of a temporal correlation between MYC2 protein accumulation and its function to differentially regulate wound response and pathogen response suggests that protein stability may play a role in MYC2 regulation . To test that MYC2 may be subjected to proteolysis in planta , we generated 35Spro:MYC2-GUS , 35Spro:MYC2-GFP and 35Spro:MYC2-4myc constructs and introduced them into the myc2-2 mutants . The resulting stable transgenic lines including MYC2-GUS-18 , MYC2-GFP-12 and MYC2-4myc-15 , which expressed comparable transcript levels of the respective transgene and rescued the JA-insensitive phenotype of the myc2-2 mutant ( Figure S1 ) , were selected for protein stability and functional analysis . Upon application of cycloheximide ( CHX ) , an inhibitor of de novo protein synthesis , GUS activity of MYC2-GUS-18 seedlings ( Figure 3A ) or GFP fluorescence of MYC2-GFP-12 seedlings ( Figure 3B ) were largely reduced . In line with a previous observation that CHX up-regulates MYC2 transcripts [28] , we showed that the mRNA levels of the MYC2-GUS or MYC2-GFP transgenes were actually increased in CHX-treated transgenic seedlings ( Figure S3 ) . These results eliminate the effect of transcriptional regulation on the protein abundance of the MYC2-GUS or MYC2-GFP fusions and support that the MYC2-reporter fusion proteins are unstable . Addition of the proteasome inhibitor MG132 to the transgenic seedlings , which barely affects the mRNA levels of the transgenes ( Figure S3 ) , lead to increased signal intensity of GUS staining ( Figure 3A ) or GFP fluorescence ( Figure 3B ) . Furthermore , co-treatment with MG132 and CHX largely blocked the effect of CHX ( Figure 3A and 3B ) . These results indicate that degradation of the MYC2-GUS or MYC2-GFP fusions requires the proteasome activity . Similarly , protein gel blot assays using MYC2-4myc-15 seedlings indicated that , whereas addition of CHX led to reduced MYC2-4myc levels , addition of MG132 led to increased MYC2-4myc levels and , the CHX effect was sufficiently suppressed by MG132 ( Figure 3C ) . As ubiquitination is a prerequisite for protein degradation by the 26S proteasome , we asked whether we could detect the ubiquitinated form of MYC2 . For this experiment , the 35Spro:MYC2-4myc construct was transferred into N . benthamiana leaves with the well-established agroinfiltration system [29] . Protein extracts from the agroinfiltrated leaves were immunoprecipitated with an anti-myc antibody and examined via western blotting using an anti-myc antibody . As shown in Figure 3D , in addition to the band of expected size for the MYC2-4myc protein , a smear of bands corresponding to larger molecules , which show the feature of ubiquitinationed forms of the MYC2-4myc fusion protein , were also detected . Indeed , when the same samples were immuno-analyzed with an anti-ubiquitin antibody , the high molecular size bands could be recognized by the anti-ubiquitin antibody , confirming that these additional bands were ubiquitinated forms of the MYC2-4myc protein ( Figure 3D ) . Together , these findings led us to a conclusion that MYC2 is subjected to UPS-dependent proteolysis . To test that MYC2 degradation is a part of the JA signaling , the above-described MYC2-GFP-12 seedlings were treated with MeJA in the absence or presence of CHX . In the absence of CHX , GFP fluorescence showed a marked increase at 6 h following MeJA treatment and , MeJA-induced elevation of GFP fluorescence was dramatically decreased by the addition of CHX ( Figure 4A ) . Similarly , as revealed by western blot assays , treatment of the MYC2-4myc-15 seedlings with MeJA alone led to a strong elevation of the MYC2-4myc fusion protein and MeJA-induced increase of the fusion protein was markedly reduced by CHX ( Figure 4B ) . Considering that addition of CHX actually showed an increasing effect on MeJA-mediated increase of the mRNA levels of the MYC2-GFP or MYC2-myc transgenes ( Figure S4 ) , these results support that UPS-mediated degradation of MYC2 occurs during JA signaling . To substantiate this observation , we monitored the MeJA-mediated change of the MYC2-GFP fluorescence of the MYC2-GFP-12 seedlings in the absence or presence of the UPS inhibitor MG132 . Upon treatment with MeJA itself , GFP signal peaked at 6 h and returned to basal levels at 48 h ( Figure 4C ) . In the presence of MG132 , however , MeJA-induced increased of GFP signal was strengthened at both 6 h and 48 h , indicating that MeJA-induced fluctuations of the GFP signal were abolished by MG132 ( Figure 4C ) . This MG132 effect on MeJA-induced fluctuations of the MYC2-GFP fluorescence provides us a facile assay to investigate the mechanism underlying MYC2 degradation . Similarly , as revealed by western blot assays , MeJA-induced fluctuations of the MYC2-4myc fusion protein levels were also abolished by MG132 ( Figure 4D ) . To evaluate the effect of MYC2 degradation on its transcription activity , we examined whether MG132 affects MeJA-induced expression of LOX2 and ORA59 , marker genes of JA-induced wound and pathogen responses that were direct targets of MYC2 . In line with the notion that MYC2 positively regulates wound response , MeJA-induced expression levels of LOX2 were largely reduced by the myc2-2 mutation ( Figure 4E ) . Whereas treatment with MG132 alone resulted in undetectable MYC2-dependent induction of LOX2 expression , the MeJA-mediated induction of this gene was strongly inhibited in the presence of MG132 ( Figure 4E ) , indicating that MeJA-induced activation of LOX2 expression requires both MYC2 and the proteasome activity . Similarly , MG132 itself showed negligible effect on MYC2-dependent regulation of ORA59 expression and this proteasome inhibitor strongly weakened the MeJA-induced activation of ORA59 expression in WT plants ( Figure 4F ) . Consistent with a negative effect of MYC2 on JA-mediated induction of ORA59 expression , MeJA-induced expression levels of ORA59 were already high in the myc2-2 mutant , and addition of MG132 showed minor , if any , effect on MeJA-induced expression of ORA59 in this mutant ( Figure 4F ) . Collectively , these results demonstrate that proteasome activity is required for the MYC2 function to differentially regulate wound response and pathogen response . Our findings that the proteasome activity is important for the MYC2 function suggest that MYC2 proteolysis is tightly linked with its transcriptional activity . It is generally considered that the N-terminal part of MYC2 is important for its transcriptional activity [3] , [4] , [12] , but the transcriptional activation domain ( TAD ) of MYC2 has not been identified based on experimental studies . We used the MATCHMAKER GAL4-based Two-Hybrid System 3 ( Clontech ) to define the TAD of MYC2 . The expressed proteins in yeast strains were analyzed by immunoblot experiments ( Figure S5 ) . In these assays , we found that MYC2 has strong transcriptional activation activity whereas MYC2ΔTAD , in which amino acids from 149 to 188 of MYC2 were deleted , dose not ( Figure 5A–5B ) , indicating that the domain from amino acid 149 to 188 could be the TAD of MYC2 . In the mammalian system , it is a general phenomenon that the destruction elements ( DE ) , which are usually acidic , overlap closely with the TADs of unstable transcription factors [18] , [28] , [30] . Indeed , our sequence analysis of the MYC2 TAD region identified a 12-amino-acid element ( MYC2154–165 ) that is enriched in acidic amino acids ( Figure 5A ) . To test that this acidic domain may function as a degron of MYC2 , we generated a DE deletion construct of MYC2 and introduced it into the myc2-2 mutant . Among the resulted transgenic plants , the line MYC2ΔDE-GFP-6 was selected for further analysis . As shown in Figure S6 , MYC2ΔDE-GFP-6 plants and the above-described MYC2-GFP-12 plants showed comparable transcript levels of the respective transgenes . Whereas blocking protein synthesis with CHX strongly reduced the MYC2-GFP fusion protein in MYC2-GFP-12 plants , CHX showed minor reduction effect on the MYC2ΔDE-GFP fusion protein in MYC2ΔDE-GFP-6 plants ( Figure 5C ) , indicating that deletion of the DE indeed affects the protein stability of MYC2 . Next , through monitoring the MG132 effect on JA-induced fluctuations of GFP fluorescence as an assay , we found that , in the absence or presence of MG132 , the MeJA-induced change of GFP fluorescence was largely abolished in MYC2ΔDE-GFP-6 plants ( Figure 5D ) . In summary , deletion of the 12-amino-acid element of MYC2 produces a protein remains stable support a scenario that the DE element we identified functions as a degron of MYC2 . ChIP-PCR assays using MYC2ΔDE-GFP-6 plants indicated that deletion of the DE does not affect the binding capacity of MYC2 to the promoter of ORA59 ( Figure S7 ) . To test that the DE may affect MYC2-directed activation of wound responsive genes , we used a transient assay to compare the activation effect of MYC2 or MYC2ΔDE on the expression of LOX2pro:LUC , a reporter containing the LOX2 promoter fused with the LUC gene . Co-expression of LOX2pro:LUC with 35Spro:MYC2 led to an obvious increase of luminescence intensity ( Figure 6A–6C ) , indicating that 35Spro:MYC2 activates the expression of LOX2pro:LUC . In contrast , the 35Spro:MYC2ΔDE construct failed to activate the expression of LOX2pro:LUC ( Figure 6A–6C ) . These results support that the DE is required for the MYC2 function to activate the expression of LOX2 . To confirm this in planta , we examined JA-induced expression of LOX2 in MYC2-GFP-12 , MYC2ΔDE-GFP-6 , myc2-2 and WT plants . MeJA-induced expression levels of LOX2 in MYC2-GFP-12 plants were much higher than those in the WT ( Figure 6D ) , indicating that the 35Spro:MYC2-GFP construct rescued the JA-insensitive phenotype of myc2-2 in term of JA-induced LOX2 expression . In MYC2ΔDE-GFP-6 plants , however , MeJA-induced expression levels of LOX2 were essentially comparable to those in the myc2-2 mutant ( Figure 6D ) , indicating that the 35Spro:MYC2ΔDE-GFP construct failed to rescue the myc2-2 mutant phenotype . These results support that the DE is important for the MYC2 function to activate the transcription of wound-responsive genes . Similarly , in a transient assay of N . benthamiana leaves , we showed that DE is also required for the MYC2 function to repress the expression of ORA59pro:LUC , a reporter containing the ORA59 promoter fused with LUC ( Figure 6A–6C ) . Comparison of JA-induced expression levels of ORA59 in MYC2-GFP-12 , MYC2ΔDE-GFP-6 , myc2-2 and WT plants indicated that , the 35Spro:MYC2-GFP construct , but not the 35Spro:MYC2ΔDE-GFP construct , rescued the myc2-2 mutant phenotype in MeJA-induced ORA59 expression ( Figure 6E ) . Given that MYC2 negatively regulates pathogen response , a featured phenotype of myc2-2 is that this mutant is more resistant than its WT counterpart to the necrotrophic pathogen Botrytis cinerea [4] . Our pathogen response assays revealed that , whereas MYC2-GFP-12 plants showed a similar performance as WT plants , MYC2ΔDE-GFP-6 plants showed a similar performance as myc2-2 plants ( Figure 6F ) . Collectively , these results lead us to a conclusion that the DE and TAD of MYC2 are functionally connected . Targeting substrates to the proteasome is often regulated by post-translational modifications , such as phosphorylation . To test if MYC2 is phosphorylated , protein extracts from the MYC2-4myc-15 transgenic plants were immunoprecipitated and treated with calf intestinal alkaline phosphatase ( CIAP ) . We found that CIAP treatment led to a slightly faster migration of the MYC2-4myc fusion protein , implying that the slower migrating form of the fusion protein was phosphorylated ( Figure 7A ) . Next , we treated MYC2-4myc-15 plants without or with MeJA , extracted total protein and applied the extracts onto a column that specifically binds phosphorylated proteins . A protein gel blot assay was performed to make sure that the amount of the MYC2-4myc protein in control and MeJA-treated samples was comparable ( Figure 7B ) . Western blot assays indicated that the MYC2-4myc fusion protein could bind to the column and , importantly , MeJA treatment led to a substantial increase of proteins bound to the column ( Figure 7B ) . These results implicate that MYC2 is phosphorylated in vivo and that MYC2 phophorylation is under the regulation of the JA signal . An examination of MYC2 protein sequence revealed a cluster of four potential phosphorylation residues including Thr328 , Ser330 , Ser334 and Thr336 ( Figure 5A ) . Mass spectrometric analysis of MYC2 immunoprecipitated from the MYC2-4myc-15 seedlings revealed phosphorylation at Thr328 ( Figure 7C ) . To find out the physiological function of the confirmed and potential phosphorylation sites , phosphorylation defective forms of the MYC2-4myc fusion protein carrying serine/threonine to alanine mutations ( i . e . , MYC2T328A , MYC2S330A , MYC2S334A , MYC2T336A ) were introduced into the myc2-2 mutant under the 35S promoter . Transgenic lines MYC2T328A-4myc-23 , MYC2S330A-4myc-13 , MYC2S334A-4myc-11 and MYC2T336A-4myc-18 , in which the expression levels of the respective transgenes were comparable to those of the MYC2-4myc-10 plants ( Figure S8 ) , were selected for further analysis . JA-induced root growth inhibition assays indicated that MYC2T328A-4myc-23 , but not the rest of the transgenic lines , showed a JA-insensitive phenotype like myc2-2 ( Figure 7D ) , revealing that the T328A mutation , but not the other three mutations , affects the MYC2 protein function . These results support that Thr328 is an in vivo phosphorylation site of MYC2 . To determine whether Thr328 phophorylation affects MYC2 turnover , MYC2-4myc-10 and MYC2T328A-4myc-23 seedlings were treated with MeJA and the abundance of the fusion proteins were examined . We found that the basal and MeJA-induced accumulation levels of MYC2T328A-4myc were higher than those of the MYC2-4myc and , importantly , that the CHX-mediated reduction of the fusion protein accumulation was abolished in MYC2T328A-4myc-23 seedlings ( Figure 7E ) . Similarly , in the absence or presence of MeJA , MG132-mediated increase of the fusion protein accumulation was also abolished in MYC2T328A-4myc-23 seedlings ( Figure 7F ) . Together , our findings that the MYC2T328A mutation renders the MYC2-4myc fusion more stable support that phosphorylation of Thr328 facilitates the proteolysis of the MYC2 protein . To test whether the MYC2T328A mutation affects MYC2-dependent gene transcription , we compared MeJA-induced expression of LOX2 and ORA59 in MYC2-4myc-10 , MYC2T328A-4myc-23 , myc2-2 and WT seedlings . As shown in Figure 8A , whereas MeJA-induced expression levels of LOX2 in MYC2-4myc-10 seedlings were strongly increased than those in WT seedlings , MeJA-induced expression levels of LOX2 in MYC2T328A-4myc-23 seedlings remained comparable to those in myc2-2 seedlings . In parallel experiments , the MeJA-induced expression levels of ORA59 in MYC2-4myc-10 seedlings were substantially reduced compared to those in myc2-2 seedlings , but the MeJA-induced expression levels of ORA59 in MYC2T328A-4myc-23 seedlings were essentially comparable to those in myc2-2 seedlings ( Figure 8B ) . In B . cinerea infection assays , the performance of MYC2-4myc-10 seedlings is similar to that of the WT seedlings , whereas the performance of MYC2T328A-4myc-23 seedlings is similar to that of the myc2-2 seedlings ( Figure 8C ) . Collectively , these results support that MYC2 phosphorylation at Thr328 is functionally coupled with its action to regulate JA-responsive gene transcription .
Several lines of evidence hint the existence of protein regulation for the function of MYC2 in regulating JA-dependent plant immunity . For example , it has been shown that MYC2 is upregulated by JA at the transcription level but , transgenic Arabidopsis plants overexpressing MYC2 or its functional homolog in tomato ( Solanum lycopsicon ) did not show constant expression of defense genes without the JA signal [2] , implying that a JA-dependent posttranscriptional modification of MYC2 is required for its function . Indeed , it was recently shown that the circadian-clock component TIME FOR COFFEE ( TIC ) interacts with and negatively regulates the protein accumulation of MYC2 [31] . A prominent action mode of MYC2 is that this transcription factor differentially regulates different subsets of JA-mediated immune responses . For example , MYC2 positively regulates the expression of early wound-responsive genes , whereas negatively regulates the expression of late pathogen-responsive genes [2]–[4] , indicating that MYC2 can act both as a transcriptional activator and repressor . However , whether protein regulation of MYC2 itself is involved in its action to temporally activate or repress specific genes remains largely unknown . Therefore , one of the major challenges in understanding the action mechanisms of MYC2 is to uncover the regulation of this transcription factor at the protein level . In this study , we found a temporal correlation between the MYC2 protein accumulation and its differential effects on the expression of wound responsive genes and pathogen responsive genes: high accumulation of the MYC2 protein correlates with peaked expression of the branch of early wound-responsive genes that are positively regulated by MYC2 , whereas low accumulation of the MYC2 protein correlates with peaked expression of the branch of late pathogen-responsive genes that are negatively regulated by MYC2 ( Figure 1 ) . Clearly , the accumulation kinetics of the MYC2 protein during JA signaling facilitates its temporal activation or repression of specific subset of genes . In the context that inducible defense is an energy costly process and that plants have evolved the ability to precisely allocate limited resources in an attacker dependent manner [32] , our findings support that MYC2 protein regulation plays an important role in resource management during JA-mediated plant immunity . The existence of a temporal correlation between MYC2 protein accumulation and its function to differentially regulate wound response and pathogen response suggests that protein stability may play a role in MYC2 regulation . Indeed , we describe here that MYC2 is subjected to UPS-dependent proteolysis and demonstrate that UPS-dependent proteolysis of MYC2 is part of the JA signaling . It is well known that UPS-dependent proteolysis plays an important role in nearly every aspect of plant biology , including plant hormone signaling . In these instances UPS either degrades transcription factors to suppress transcription or degrades transcription repressors to activate gene expression [33]–[36] . Surprisingly , we provide here evidence that UPS-dependent proteolysis plays a counterintuitive role in the regulation of MYC2 during JA-mediated immune responses: First , UPS activity is essential for the MYC2 function to temporally activate or repress specific genes; Second , a 12-amino acid element in the TAD of MYC2 plays dual roles to signal proteolysis and to regulate transcription activity of MYC2 . These data support that UPS-dependent turnover of MYC2 is coupled with its transcription activity during JA-mediated plant immunity . We further show that phosphorylation at Thr328 is important for both the proteolysis and the transcription activity of MYC2 , indicating that UPS-mediated destruction of MYC2 is inherently linked to the way in which it stimulates gene transcription . In the plant kingdom , a similar regulatory mechanism was observed recently for the regulation of NONEXPRESSOR OF PATHOGENESIS-RELATED GENES ( NPR1 ) , a transcription coactivator involved in salicylic acid ( SA ) -mediated plant immunity [37] . Our finding that UPS-mediated proteolysis is involved directly and mechanistically in the regulation of MYC2 fits well with a scenario in the animal and yeast called “activation by destruction” , in which UPS-mediated proteolysis can activate the activity of the transcriptional regulators it destroys [13] , [18]–[20] , [30] , [37]–[41] . This “activation by destruction” phenomenon has only been appreciated in recent years , but it appears to apply to an ever growing number of transcriptional regulators in animal and yeast [13] , [19] , [20] . Current hypotheses to explain the connection between transcription regulator destruction and function hold three main points: First , destruction of transcription factors is likely to be a direct consequence of their ability to activate transcription; Second , turnover of transcription factors occurs on the chromatin and during the process of gene induction; Third , turnover of transcription factors is often signaled by kinases that are integral parts of the transcriptional machinery [13] , [38] . Based on these hypotheses and our findings in this study , we propose a model to explain how the UPS-dependent proteolysis stimulates the transcription activity of MYC2 . On the chromatin and during JA-mediated induction process of MYC2 target genes , kinases associated with the transcriptional machinery serve to mark MYC2 as “spent” , trapping it in an inactive state . At the same time , these phosphorylation events bring the UPS machinery in and therefore destroy the “spent” MYC2 in situ , clearing the deck for promoter association with a “fresh” MYC2 molecule . In this model , the initial ‘pioneer round’ of transcription does not involve UPS-dependent proteolysis of MYC2 . Proteolysis plays a positive role in transcription by allowing “fresh” MYC2 to access the promoter and therefore stimulates additional rounds of transcription . It will be interesting in future studies to prove that phosphorylation-coupled proteolysis of MYC2 occurs on the chromatin and to identify the kinases and E3 ligases involved in the phophorylation and proteolysis of MYC2 . Regardless of these open questions , this study clearly demonstrates that phosphorylation-coupled proteolysis of transcription factors may be a common mechanism by which higher plants regulate gene transcription .
Arabidopsis thaliana ecotype Columbia ( Col-0 ) was used as wild type ( WT ) . T-DNA insertion mutant myc2-2 was previously described [1] , [27] . Arabidopsis plants were grown in Murashige and Skoog ( MS ) media at 22°C with a 16-h-light/8-h-dark photoperiod ( light intensity 120 µM photons m−2s−1 ) as previously described [42] . For MeJA-induced gene expression and protein accumulation assays , seedlings of indicated ages grown under 16-h light/8-h dark were treated with 100 µM MeJA . MeJA-treated seedlings were then transferred to continuous light for indicated times . The 35Spro:MYC2-GUS was prepared by inserting PCR-amplified coding sequence of MYC2 and Glucuronidase ( GUS ) into the KpnI-SpeI sites and SalI-PstI sites of the binary vector pCAMBIA2300 under the control of cauliflower mosaic virus 35S promoter . The 35Spro:MYC2-GFP and 35Spro:MYC2-4myc constructs used in this study were previously described [1] , [27] . To generate 35Spro:MYC2ΔDE-GFP construct , coding sequence of MYC2 truncated DE domain was amplified with Gateway-compatible primers . The PCR product was cloned by pENTR Directional TOPO cloning kit ( Invitrogen ) and then recombined with the binary vector pGWB5 ( 35S promoter , C-GFP ) [43] . The codon for Thr328 of MYC2 in pENTR-MYC2 [1] was replaced with the amino acid encoding alanine using the TaKaRa MutanBEST kit . The mutation was confirmed by DNA sequencing . The pENTR-MYC2T328A was then combined with binary vector pGWB17 ( 35S promoter , C-4myc ) [43] to generate 35Spro:MYC2T328A-4myc construct . Similarly , we generated 35Spro:MYC2S330A-4myc , 35Spro:MYC2S334A-4myc , 35Spro:MYC2T336A-4myc constructs . All primers used for DNA construct generation are listed in Table S1 . The above constructs were then transformed into Agrobacterium tumefaciens strain GV3101 ( pMP90 ) , which was used for transformation of Arabidopsis plants by vacuum infiltration [44] . The transient expression assays were performed in N . benthamiana leaves as previously described [1] , [45] . The ORA59 promoter was amplified cloned into pENTR using the pENTR Directional TOPO cloning kit ( Invirogen ) . To generate ORA59 promoter with mutations , site-directed mutagenesis was used to delete the CACGTG motif in the P1 region of the ORA59 promoter ( Figure 2A ) using the TaKaRa MutanBEST kit . Similarly , the LOX2 promoter was amplified and cloned into pENTR vector . Then various promoter versions were fused with the luciferase reporter gene LUC through the Gateway reactions into the plant binary vector pGWB35 [43] to generate the reporter constructs ORA59pro:LUC , ORA59mpro:LUC , LOX2pro:LUC . The MYC2 and MYC2ΔDE effector constructs were the above-described 35Spro:MYC2-GFP ( 35Spro:MYC2 ) , and 35Spro:MYC2ΔDE-GFP ( 35Spro:MYC2ΔDE ) . We used a low-light cooled CCD imaging apparatus ( NightOWL II LB983 with indigo software ) to capture the LUC image and to count luminescence intensity . The leaves were sprayed with 100 mM luciferin and were placed in darkness for 3 min before luminescence detection . Full-length coding sequence and TAD truncation of MYC2 were amplified with listed primers ( see Table S1 online ) . Enzyme-digested PCR products were cloned into the NdeI and PstI sites of the vector pGBKT7 . The resulting constructs were then transformed into the yeast strain Saccharomyces cerevisiae AH109 . The MATCHMAKER GAL4-based Two-Hybrid System 3 ( Clontech ) was used for the transactivation activity assay . Each yeast liquid culture was serially diluted to OD600 = 0 . 6 , and 5 µl of each dilution was inoculated onto SD/-Ade/-His/-Trp synthetic dropout medium . The expressed proteins in yeast strains were analyzed by immunoblot experiments . Proteins fused with the GAL4 DNA binding domain were detected using anti-myc antibody . For qRT-PCR analysis , seedling were harvested and frozen in liquid nitrogen for RNA extraction . RNA extraction and qRT-PCR analysis were performed as previously described [27] . Primers used to quantify gene expression levels are listed in Table S2 . Details for protein extraction and immunoblot assays were described recently [27] . Antibodies and dilutions used in these experiments were as follows: anti-myc antibody ( Abmart , 1∶2000 ) , anti-GFP antibody ( Abmart , 1∶1000 ) . anti-ubiquitin antibody ( Sigma , 1∶1000 ) . N . benthamiana leaves that transiently expressing MYC2-4myc or MYC2-4myc-15 seedlings were homogenized with ice-cold extraction buffer ( 50 mM Tris-HCl , pH 7 . 5 , 150 mM NaCl , 0 . 1% Triton X-100 , 0 . 2% Nonidet P-40 , 0 . 6 mM PMSF , and 20 µM MG132 with Roche protease inhibitor cocktail ) . After protein extraction , 20 µL protein G plus agarose ( Santa Cruz ) was added to the 2-mg extracts to reduce nonspecific immunoglobulin binding . After 1 h of incubation , the supernatant was transferred to a new tube . myc antibody-bound agarose beads ( Santa Cruz ) were then added to each reaction for 4 h at 4°C . The precipitated samples were washed at least four times with the extraction buffer and then eluted by adding 1× SDS protein loading buffer with boiling for 5 min . A phosphor-portein enrichment kit ( Clontech ) was used to column-purify phospho-proteins according to the manufacturer's protocol . ChIP was performed as previously described [46] . 1 . 5 gram of 10-d-old 35Spro:MYC2-GFP seedlings were used for ChIP experiments . GFP antibody ( Abcam ) was used to immunoprecipitate the protein-DNA complex . Chromatin precipitated without antibody was used as negative control , while the isolated chromatin before precipitation was used as input control . The enrichment of DNA fragments was determined by semiquantitative PCR . Primers used for ChIP-PCR were listed in Table S2 . Recombinant MYC2 protein in Escheichia coli ( E . coli ) used in this assay was previously described [1] . Oligonucleotide probes were synthesized and labeled with biotin at the 3′ end ( Invitrogen ) . EMSA was performed using a Lightshift Chemiluminescent EMSA Kit ( Thermo Scientific ) . Briefly , biotin-labeled probes were incubated in 1× binding buffer , 2 . 5% glycerol , 50 mM KCl , 5 mM MgCl2 and 10 mM EDTA with or without proteins at room temperature for 20 min . For nonlabeled probe competition , nonlabeled probe was added to the reactions . The probe sequences used in MESA were listed in Table S3 . The anti-myc immunoprecitipitates from MYC2-4myc-15 transgenic plants were resolved on SDS-PAGE and visualized by silver staining . A protein band of approximately 70 KD was cut from the gel and digested with trypsin overnight at 37°C . Peptides were extracted sequentially with 5% formic acid ( FA ) /50% acetonitrile and 0 . 1% FA/75% acetonitrile , dried under vacuum , then resuspended in 0 . 1% FA . LC-MS/MS and data analysis were performed as described previously [47] except that the peptide sample was loaded directly on an analytical reverse-phase column . Grow Botrytis cinerea on MEA medium ( 2% malt extract , 2% glucose , 0 . 1% peptone , 2% agar ) in Petri dishes for 14 days at 24°C with 12 h photoperiod before collection of spores . Spore inoculums were prepared by harvesting spores in water , filtration though nylon mesh to remove hyphae and suspension in potato dextrose broth to a concentration of 105 spores/ml . Detached rosette leaves of 28-d-old plants were placed in Petri dishes containing 0 . 8% agar , with the petiole embedded in medium . Each leaf was inoculated with a single 5 µl droplet of B . cinerea inoculums . Trays were covered with lids and kept under the same conditions as for plant growth . Photographs were taken after 3 days and mean lesion sizes of 20 leaves from 20 plants of various genotypes were compared using a Student's t-test assuming equal variance .
|
The plant hormone jasmonic acid ( JA ) regulates a wide range of plant immune responses involving genome-wide transcriptional reprogramming that are regulated by the basic helix-loop-helix ( bHLH ) Leu zipper transcription factor MYC2 . As a master regulator of JA signaling , MYC2 differentially regulates the transcription of different branches of JA–responsive genes through distinct molecular mechanisms . Here , we provide evidence that phosphorylation-dependent turnover of MYC2 is coupled with its function . We show that , during JA response , high accumulation of the MYC2 protein correlates with peaked expression of early wound-responsive genes that are positively regulated by MYC2 , whereas low accumulation of the MYC2 protein correlates with peaked expression of late pathogen-responsive genes that are negatively regulated by MYC2 . We discovered a 12-amino-acid element in the transcription activation domain of MYC2 that is required for both the proteolysis and the transcriptional activity of MYC2 . Interestingly , MYC2 phosphorylation at residue Thr328 , which facilitates its turnover , is also important for the MYC2 function to regulate transcription . Together , these results reveal that phosphorylation and turnover of MYC2 are tightly linked with its function to regulate the transcription of JA–responsive genes . Our results exemplify that plants employ proteolysis-coupled transcription as mechanism to fine-tune their responses to versatile stresses .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biology"
] |
2013
|
Phosphorylation-Coupled Proteolysis of the Transcription Factor MYC2 Is Important for Jasmonate-Signaled Plant Immunity
|
Theoretical advances in the science of consciousness have proposed that it is concomitant with balanced cortical integration and differentiation , enabled by efficient networks of information transfer across multiple scales . Here , we apply graph theory to compare key signatures of such networks in high-density electroencephalographic data from 32 patients with chronic disorders of consciousness , against normative data from healthy controls . Based on connectivity within canonical frequency bands , we found that patient networks had reduced local and global efficiency , and fewer hubs in the alpha band . We devised a novel topographical metric , termed modular span , which showed that the alpha network modules in patients were also spatially circumscribed , lacking the structured long-distance interactions commonly observed in the healthy controls . Importantly however , these differences between graph-theoretic metrics were partially reversed in delta and theta band networks , which were also significantly more similar to each other in patients than controls . Going further , we found that metrics of alpha network efficiency also correlated with the degree of behavioural awareness . Intriguingly , some patients in behaviourally unresponsive vegetative states who demonstrated evidence of covert awareness with functional neuroimaging stood out from this trend: they had alpha networks that were remarkably well preserved and similar to those observed in the controls . Taken together , our findings inform current understanding of disorders of consciousness by highlighting the distinctive brain networks that characterise them . In the significant minority of vegetative patients who follow commands in neuroimaging tests , they point to putative network mechanisms that could support cognitive function and consciousness despite profound behavioural impairment .
There has been considerable recent interest in the view that consciousness is a phenomenon emerging from the dynamic equilibrium between differentiated and integrated information processing in the brain [1]–[4] . This view has inspired research into ways of quantifying the characteristics of information exchange in the brain at rest , and how this modulated in natural sleep , pharmacological sedation , and pathological coma and disorders of consciousness ( DoC; including the vegetative and minimally conscious states , VS and MCS ) . In this latter case , such theoretical questions about the neural bases of consciousness take on a clinical and societal significance , as they could inform diagnosis , prognosis and treatment of DoC , which are often brought on by severe injury to the brain . Recent advances in the use of neuroimaging to better ascertain brain function in DoC have yielded some surprises , and indicated that a significant minority of patients are able to volitionally modulate brain activity in ways that would normally require high-level cognition and even covert awareness despite no behaviourally evident signs thereof [5]–[14] . Such findings have motivated parallel research into the study of brain connectivity in patients at rest , using MRI [15] , [16] , EEG [17]–[19] and TMS [20] , [21] to derive surrogate measure of information integration and differentiation . Modern neuroimaging methods for assaying such connectivity , including Magnetic Resonance Imaging ( MRI ) and high-density electroencephalography ( EEG ) , provide a surfeit of data that need to be reduced in dimensionality and coalesced into patterns to provide an overarching understanding of connectivity networks in the brain . Graph-theoretical analysis of such networks [22]–[24] has provided an elegant way to achieve this synthesis using resting state connectivity data [23]–[27] in sleep [28]–[31] , sedation [32] and coma [33] . Here , we apply graph theory to extract patterns of information integration in brain networks derived from bedside measurement of high-density EEG in DoC patients , alongside normative networks observed in healthy controls . From 10 minutes of high-density EEG data , we calculate networks of sustained , coherent oscillatory activity within canonical frequency bands , which are prominent and commonly clinically evaluated in DoC . We will show that graph-theoretical metrics highlight contrasting signatures of connectivity in healthy and pathological brains across different frequency bands . These signatures , encompassing measures of topology as well as topography , will allow us to address a set of inter-related questions of fundamental neuroscientific importance: for example , what is distinctive about network dysfunction in pathological states of low awareness ? To what extent are these network signatures consistent across patients ? How do they correlate with the complexity of preserved behavioural responses ? And perhaps most intriguingly , what network signatures can we observe in patients who seem behaviourally vegetative , but nevertheless demonstrate signs of covert awareness .
Figure 1A plots the log spectrum of each channel , averaged across 26 healthy controls . It generally conforms to the 1/f ‘pink noise’ decay that characterises human EEG , and is punctuated by prominent peaks in the canonical delta ( 0–4 Hz ) and alpha ( 8–13 Hz ) frequency bands . The topographic contributions of spectral power within these bands ( see supplementary figure S1 ) shows that the power of the delta band peak is relatively more concentrated in frontal electrodes whereas the alpha peak is prominent in bilateral occipital electrodes . Across the spectra in the 3 groups in figures 1A–C , there was a prominent drop-off in alpha power in both MCS and VS patients , with a corresponding increase in delta power . This overall ‘slowing-down’ of resting state EEG , or slow-wave activity after severe brain injury has been documented [34] . To quantify this , we measured the channel-wise contribution to the average power across all channels within 0–40 Hz , from each frequency band . As expected , there were statistically significant differences in the relative power contributions from the delta , alpha and beta bands: as shown in figure 1D , patients together generated significantly more power in the delta band than healthy controls ( Unequal variances t ( 55 . 7 ) = 9 . 97 , p<0 . 001 ) . In fact , 80% of overall spectral power in VS patients was concentrated within the delta band . The reverse was true in the alpha and beta bands: patients had significantly smaller power contributions from the alpha band ( t ( 32 . 3 ) = 10 . 0 , p<0 . 001 ) . There was no significant difference in the power contribution from the theta band across the two groups . Visually , patient spectra in figures 1B and C suggested broadband increases in power in the higher i . e . , beta and gamma frequency bands , due to elevated levels of electromyographic ( EMG ) noise introduced by involuntary muscle movements , especially in MCS patients . Hence , to avoid the consequent potential confounds in comparisons of activity in these bands between patients and healthy controls , we restricted ourselves to further analysis of data only within the alpha , theta and delta bands , where there were primary effects of interest and the influence of EMG noise was negligible . While increased delta and theta power can sometimes be attributed to sleep-onset related EEG activity , we conventionally followed a protocol to ensure that our patients were arousable at the beginning of data acquisition ( see Materials and Methods for details ) , which made it unlikely that they were consistently asleep during the following 10-minute period . We confirmed this by calculating the amount of temporal variability in power contributions within each frequency band in controls and patients . If patients were indeed falling asleep during the data acquisition , we would expect to see relatively higher variation in their delta band , alongside a progressive reduction in eye-movement related behaviour over the duration of the 10-minute recording . However , as can be seen in supplementary figure S2 , there was no evidence of such higher variability: we found that healthy controls had higher temporal variations in delta ( t ( 42 . 4 ) = 3 . 16 , p = 0 . 003 ) and alpha ( t ( 36 . 0 ) = 12 . 69 , p<0 . 001 ) power . Further , we observed no consistently progressive decline in eye-movement related behaviour in patients , as recorded by derived electrooculographic ( EOG ) channels . As supplementary figure S3 depicts , though there was considerable variability in EOG activity between patients and over time , average activity in the first half of the recording session was statistically indistinguishable from the second half across patients ( t ( 31 ) = 1 . 17 , p = 0 . 25 ) . This lack of a systematic decline suggested the absence of an unequivocal indication of sleep onset in the patient group . While oscillatory power in the lower frequency bands derives from the predominance of so-called slow wave activity , alpha oscillations have been linked to arousal , attention and alertness [35] . Having observed significant group-level differences in power between patients and controls in these bands , we investigated the link between delta and alpha oscillations and clinically evidenced arousal in our patient group by quantifying the extent to which alpha power contributions could explain scores on the Coma Recovery Scale-Revised ( CRS-R ) . The results of robust multilinear regression of delta and alpha power contributions as predictors of CRS-R scores are shown in figures 1E and F . As the regression lines plotted therein depict , there was a statistically significant link between a decrease in delta power , and a complementary increase in alpha power , with increase in CRS-R scores . This finding replicates the pattern recently reported by Lechinger et al . [36] , who demonstrated a similar link between alpha power ( and peak frequency ) and CRS-R scores . In our data this trend was particularly visible in the MCS patient group ( indicated in blue in figures 1E and F ) . These findings , taken together with previous evidence , are convergent with the notion that the presence of fast cortical oscillations is correlated with behavioural function in DoC . We assessed connectivity between EEG electrodes to investigate the structure of brain networks in the delta , theta and alpha bands . The extent of spectral coherence between every pair of electrodes was calculated using the debiased weighted Phase Lag Index ( dwPLI ) metric [37] . dwPLI is a sensitive measure of true connectivity between cortical regions that has been shown to be robust against the influence of volume conduction , uncorrelated noise , and inter-subject variations in sample size . An earlier incarnation of this measure , the Phase Lag Index [PLI , see 38] , has been applied to low-density EEG acquired from DoC patients at rest , to show that those in VS elicit lower PLI values than MCS [18] . Here we calculated dwPLI for 91 channels in a high-density mesh ( see Materials and Methods for details ) , to investigate the latent structure of connectivity networks that emerged within and across these groups . Figures 2A–C depicts topographic maps of dwPLI-derived connectivity for each group of subjects in each of the frequency bands of interest , visualising the opposing patterns of structure observed therein . To plot these maps , 91×91 dwPLI connectivity matrices were averaged within each group and plotted as topographs with electrodes as nodes and dwPLI values as edges , and graph-theoretical algorithms were employed to automatically identify modular structure therein . The topological structure of networks of alpha band connectivity in controls ( see figure 2C , left ) highlighted the presence of prominent modules ( differentiated by colour ) consisting of dense long-range synchrony that linked occipital , parietal and frontal electrodes . This structure was distinct from the bilateral , occipitally centered , distribution of alpha power of the scalp ( see supplementary figure S1C , left ) , reinforcing the notion that dwPLI was measuring connectivity distinct from the effects of local volume conduction . It was also convergent with the notion that alpha networks observed over the healthy brain reflect broadly synchronous ambient cortical rhythms that are coeval with arousal and alertness . Such long-range connectivity structure in alpha connectivity was clearly lacking in patients , as is visually evident in figure 2C ( middle and right ) where a predominance of spatially localised , short-range synchrony was observed . Indeed , connectivity was generally weaker in patients , with mean dwPLI over all channel pairs significantly higher in controls ( t ( 51 . 7 ) = 5 . 22 , p<0 . 001 ) . Interestingly however , a contrasting pattern was evident in the delta and theta network topologies of the 3 groups , plotted in figures 3A and 3B: VS and MCS patients appeared to have relatively robust connectivity in the delta and theta bands . In particular , we observed the presence of prominent meso-scale modules in both VS and MCS patients . This suggested that the previously noted presence of higher power in the lower frequency bands in patients was concomitant with structured patterns of connectivity . Mean dwPLI of patients was also higher in the delta ( t ( 40 . 3 ) = 2 . 82 , p = 0 . 007 ) and theta ( t ( 43 . 0 ) = 2 . 93 , p = 0 . 005 ) bands . To quantify these structural differences in connectivity , we calculated and compared graph-theoretical summary measures of the topographs . The subject-wise dwPLI connectivity matrices in each band were thresholded at varying levels of connection density , to retain between 50–10% ( step size of 2 . 5% ) of the strongest dwPLI values in each matrix . At each value of this connection density threshold , we calculated 4 commonly measured metrics derived from graph theory , which captured topological properties of the networks observed in the thresholded connectivity matrices . Figure 3 plots the clustering coefficient , characteristic path length , modularity , and participation coefficient in each band and group , averaged across all connection densities considered . Supplementary plots the trends in these metrics as a function of decreasing connection density . These trends were generally consistent across the range of densities considered . Having established that there were consistent differences between EEG resting networks observed in patients and healthy controls , we set out to assess the link between graph-theoretic metrics derived from these networks and clinical evaluations of neurological and cognitive function in individual patients . As figures 3 , 4 and 5 suggest , there were no prominent statistically significant differences between the metrics obtained in VS and MCS patients , potentially attributable to the nature of our convenience sample ( see Discussion section ) . Focusing instead on the CRS-R scores of the patients , we noted that Lechinger et al . [36] recently showed a correlation between power ratios and peak frequency in the alpha band and CRS-R scores , a finding similarly observed in our data ( see figure 1F ) . Following on from this , we first investigated whether mean dwPLI values of patient connectivity matrices were correlated with CRS-R scores . No significant correlation was found in any of the frequency bands of interest . Going beyond assessment of mean spectral power and connectivity , figures 6A–D plot the robust linear regressions of key graph-theoretic metrics of patient alpha networks , which significantly predicted CRS-R scores . Note that figure 6C plots average NMI between the alpha networks of a patient and each control ( encompassing between-group values within the magenta boxes in figure 4C ) , indexing the degree to which a patient's network structure was like that of healthy brains . There was a common pattern in the variation of these metrics that separated VS from MCS patients ( plotted as red and blue circles , respectively ) . VS patients in our dataset , who by definition did not present much behavioural variation and were assigned CRS-R scores of either 7 or 8 , nevertheless had considerable variation in the efficacy of their alpha band networks as measured by graph theoretical analysis . We found that some of this variation could be interpreted in light of independent evidence that some of the VS patients considered performed tennis imagery detected by fMRI ( see Materials and Methods for details; also see Owen et al . [5] and Monti et al . [9] ) . The 4 VS patients who performed tennis imagery , indicated by filled circles in figures 6A–D , tended to be outliers in terms of the characteristics of their alpha band networks . As a representative example , the networks of three patients P1 , P2 and P3 , all with CRS-R scores of 7 but of whom only P3 performed tennis imagery , are shown in figures 6E , F and G . As is evident , alpha band connectivity in P3 was comparatively much better preserved , and was clearly distinct from the lack of structure in P2 or P3 . Specifically , P3's alpha network , unlike P1's or P2's , was remarkably similar to healthy controls , and consisted of strong , long-range connections spanning occipital , parietal and frontal regions ( compare figures 6E , F and G with figure 2C , left ) . In keeping with this visual interpretation , quantitative graph-theoretic metrics of P3's alpha band network , highlighted in figures 6A–D , were also exceptional compared to P1 and P2 , and the rest of the VS patients . In particular , P3's alpha network had much higher local connectivity ( figure 6A ) , more inter-modular hub nodes ( figure 6B ) , and structurally similar modules to healthy controls ( as measured by relative-to-healthy NMI , see figure 6C ) that spanned greater topographical distances ( as measured by modular span , see figure 6D ) . In particular , though these prominent differences between the brain networks of P1 and P3 could perhaps be attributed to aetiology , it could not explain away the differences between P2 and P3 , as both had suffered traumatic brain injury . It is also interesting to note that though P3's alpha network properties were clearly very prominent outliers as compared to P1 and P2 , delta and alpha power in P3 were much less exceptional ( compare figures 6A–D to figures 1E and F ) . Hence characterising network signatures of spectral connectivity could considerably improve our understanding of residual brain function in behaviourally uncommunicative patients who nevertheless demonstrate covert awareness . Among MCS patients , in whom there was more meaningful variability in CRS-R scores , we observed a clear trend toward increasing clustering , network centrality and modular span of alpha band networks as CRS-R scores improved ( dashed regression lines in figures 6A–D ) . In other words , as the alpha networks of MCS patients approached levels of structured , long-range , inter-modular connectivity seen in healthy controls , their CRS-R scores got progressively higher . It is worth noting that the same regressions when including all patients ( indicated by the solid lines ) were much weaker due to the considerable variation of network metrics within the VS group in the absence of matching behavioural variation . We found that there were no significant correlations between graph metrics calculated from delta and theta band networks and CRS-R scores of patients , despite the significantly higher levels of topological connectivity observed in these networks ( see figures 3 and 4 ) . As we discuss next , this could potentially be explained by the limited topographical extent of connectivity in lower frequency bands . Overall , our results suggested that it was the structure of connectivity in the alpha band that generated the strongest link to behaviourally evidenced neurological and cognitive function in patients . Finally , we examined whether there were systematic differences between graph-theoretic metrics of patients who evidenced covert awareness as measured by tennis imagery , vs . patients who did not . While none of the metrics of patients with positive imagery ( filled circles in Figure 6 ) significantly differed from those of patients without such evidence , a key distinction was evident between the VS and MCS subgroups: as pointed out earlier , amongst the VS patients , those with evidence of positive imagery tended to have remarkably higher alpha graph-theoretic metrics . This difference could not be statistically verified due to the lack of power ( only 4 out of 13 VS patients performed imagery ) . However , the pattern was reversed in MCS patients , where those with positive imagery tended to exhibit lower values of alpha metrics . In particular , inter-modular centrality ( as measured by the SD of participation coefficients ) was significantly lower in MCS patients who performed tennis imagery ( t ( 10 . 8 ) = 2 . 93 , p = 0 . 014 ) , though modular spans ( t ( 10 . 2 ) = 2 . 27 , p = 0 . 046 ) , clustering coefficients ( t ( 9 . 5 ) = 2 . 19 , p = 0 . 054 ) and characteristic path lengths ( t ( 13 . 8 ) = 1 . 88 , p = 0 . 081 ) were not . This somewhat paradoxical finding could potentially be explained by the observation ( see figure 6 ) that MCS patients who did not perform tennis imagery also tended to score higher on the CRS-R scale , though their scores were not significantly higher . Hence it could be that these patients had progressed into a post-traumatic confusional state with potentially limited attention control , known to characterise emergence from MCS [14] , .
Our analysis of EEG connectivity in high-density networks at rest found that DoC patients had comparatively reduced graph-theoretic network efficiency in the alpha band as compared to healthy controls . Using a novel metric termed modular span that embedded topologically derived modules in topographical space , we established that the alpha network modules in patients were also spatially limited , with a prominent absence of the structured long-distance connectivity commonly observed in healthy networks . Importantly however , the observed differences between graph-theoretic metrics were partially reversed in the networks within the delta and theta bands . Here we noted the presence of robust connectivity patterns that were in fact commonly structured across patients , suggesting that there could be some degree of reorganisation , rather than just disorganisation , of brain networks in DoC . However , network modules in these lower bands did not have spatial spans that characterised healthy alpha modules . This finding addresses the question of why these lower band networks could not subserve balanced cortical integration and differentiation thought to be concomitant with normal consciousness . Going further , we found that alpha network metrics in patients clearly correlated with their behavioural scores on the CRS-R . Interestingly , we observed that some behaviourally vegetative patients who demonstrated evidence of command following with fMRI tennis imagery tended to deviate from this trend: their alpha networks were remarkably well preserved and were similar to those observed in healthy controls . On the whole , our findings describe distinctive signatures of brain networks in chronic disorders of consciousness . Further , in the significant minority of vegetative patients who show signs of covert awareness , they point to putative network mechanisms that could support high-level cognitive function despite behavioural impairment .
All healthy controls gave written informed consent . Ethical approval for testing healthy controls was provided by the Cambridge Psychology Research Ethics Committee ( CPREC reference 2009 . 69 ) and the institutional ethics committee of the Faculty of Psychology of Universidad Diego Portales . Written informed consent was acquired from all patients' families and medical teams . Ethical approval for testing patients was provided by the National Research Ethics Service ( National Health Service , UK; LREC reference 99/391 ) . All clinical investigations were conducted in accordance with the Declaration of Helsinki . From each participant , we collected at least 10 minutes of 128-channel high-density EEG data in microvolts ( uV ) , sampled at 250 Hz and referenced to the vertex , using the Net Amps 300 amplifier ( Electrical Geodesics Inc . , Oregon , USA ) . Resting state data from healthy controls were acquired in a state of relaxed eyes-open wakefulness , while fixating on a central cross to minimise eye movements . Eye-blink activity was visually evaluated to ensure that the controls had their eyes open throughout the 10-minute recording . Data from patients was acquired with a consistent protocol that was conventionally employed to ensure that the patient had eyes open and was aroused at the beginning of data collection . In addition , data was collected with most patients in a sitting position , unless clinical circumstances necessitated otherwise , as previous research has shown that the supine position adversely affects arousal and behavioural responsiveness [52] . To objectively assess eyes-open/eyes-closed states , we measured eye-blink and eye-movement related activity in our data . To this end we derived left and right vertical bipolar electrooculographic ( EOG ) channels from our raw EEG data , as subtractions of channels 25 vs . 127 , and 8 vs . 126 , respectively . Similar to the approach employed by Cologan et al . [53] we filtered these derived channels with 1–3 Hz to focus on eye-movement related activity , and then calculated their standard deviations ( SD ) within a 1-second non-overlapping sliding window over time , normalised by the average SD over all such windows . Supplementary figure S3 plots the time course of this normalised SD for each patient , averaged over these two bipolar channels . Data from 91 channels over the scalp surface ( at locations shown in Figure 7 , top left ) were retained for further analysis . Channels on the neck , cheeks and forehead , which mostly contributed more movement-related noise than signal in patients , were excluded . Exactly 10 minutes of continuous data were retained , filtered between 0 . 5–45 Hz , and segmented into 60 10-second long epochs . Each epoch thus generated was baseline-corrected relative to the mean voltage over the entire epoch . Data containing excessive eye movement or muscular artefact were rejected by a quasi-automated procedure: abnormally noisy channels and epochs were identified by calculating their normalised variance and then manually rejected or retained by visual inspection . Independent Components Analysis ( ICA ) based on the Infomax ICA algorithm [54] was used to visually identify and reject noisy components . After pre-processing , a mean ( SD ) of 54 ( 7 ) , 53 ( 7 ) , 55 ( 2 ) epochs were retained for further analysis in VS , MCS patients and healthy controls , respectively . An ANOVA revealed no statistically significant difference between the numbers of epochs retained in the groups . Finally , previously rejected channels were interpolated using spherical spline interpolation , and data were re-referenced to the average of all channels . These processing steps were implemented using custom MATLAB scripts based on EEGLAB [55] . Figure 7 depicts the data processing pipeline employed to calculate spectral power and connectivity measures from the clean EEG datasets . Spectral power values within bins of 0 . 25 Hz were calculated using Fourier decomposition of data epochs using the pwelch method . At each channel , power values within five canonical frequency bands , delta ( 0–4 Hz ) , theta ( 4–8 Hz ) , alpha ( 8–13 Hz ) , beta ( 13–30 Hz ) and gamma ( 30–40 Hz ) were converted to relative percentage contributions to the total power over all five bands . Alongside , cross-spectrum between the time-frequency decompositions ( at frequency bins of 0 . 49 Hz and time bins of 0 . 04 s ) of every pair of channels was used to calculate a debiased , weighted Phase Lag Index ( dwPLI ) as introduced by Vinck et al . [37] . Generally speaking , phase synchronisation , widely seen as an EEG measure of information exchange between neuronal populations , is often calculated from the phase or the imaginary component of the complex cross-spectrum between the signals measured at a pair of channels . For example , the well-known Phase Locking Value ( PLV; see Lachaux et al . [56] ) is obtained by averaging the exponential magnitude of the imaginary component of the cross-spectrum . But many such phase coherence indices derived from EEG data are affected by the problem of volume conduction [57] , [58] , as a result of which a single dipolar source , rather than a pair of distinct interacting sources , can produce spurious coherence between spatially disparate EEG channels . The Phase Lag Index ( PLI ) , first proposed by Stam et al . [38] attempts to minimise the impact of volume conduction and common sources inherent in EEG data , by averaging the signs of phase differences , thereby ignoring average phase differences of 0 or 180 degrees . This is based on the rationale that such phase differences are likely to be generated by volume conduction of single dipolar sources . But despite being insensitive to volume conduction , PLI has two important limitations: firstly , there is a strong discontinuity in the measure , which causes it to be maximally sensitive to noise; secondly , when calculated on small samples , PLI is biased towards strong coherences ( i . e . , it has a positive sample-size bias ) . The Weighted PLI measure ( wPLI; see Vinck et al . [37] ) addresses the former problem by weighting the signs of the imaginary components by their absolute magnitudes . The Debiased Weighted PLI ( dwPLI ) additionally addresses the latter problem by being minimally biased when the number of epochs is small . Further , as the calculation of wPLI also normalises the weighted sum of signs of the imaginary components by the average of their absolute magnitudes , it represents a dimensionless measure of connectivity that is not directly influenced by differences in spectral or cross-spectral power . For these reasons , we employed the dwPLI measure to estimate connectivity in our data . For a particular channel pair and frequency band , the peak dwPLI across all time and frequency bins within that frequency band was recorded as the ambient amount of connectivity between those channels . Due to relatively higher levels of noise due to muscular artefact observed in patient spectra ( see figure 1 ) , this calculation of dwPLI-derived connectivity was restricted to the delta , alpha and theta bands , where the impact of such noise relatively negligible , and prominent differences between the power spectra were observed . The 91×91 subject-wise , band-wise dwPLI connectivity matrices thus estimated were thresholded to retain between 50–10% of the largest dwPLI values . They were then represented as graphs with the electrodes as nodes and non-zero values as edges . The lowest threshold of 10% ensured that the average degree was not smaller than , where N is the number of nodes in the network ( i . e . , N = 91 ) . This lower boundary guaranteed that the resulting networks were estimable [39] . Similar ranges of graph connection densities have been shown to be the most sensitive to the estimation of ‘true’ topological structure therein [33] , [59]: higher levels of connection density result in increasingly random graphs , while lower levels result in increasingly fragmented graphs . At each step of the connection density between 50% and 10% in steps of 2 . 5% , the thresholded graphs were submitted to graph-theoretical algorithms implemented in the Brain Connectivity Toolbox [60] . These algorithms were employed to calculate metrics that captured key topological characteristics of the graphs at multiple scales . These included the micro-scale clustering coefficient and macro-scale characteristic path length [39] and global efficiency [41] , alongside meso-scale measures like modularity and community structure [using the Louvain algorithm , see 61] , and participation coefficient [43] . Modularity and community structure calculated by the heuristic Louvain algorithm , and all measures derived therefrom , were averaged over 50 repetitions . In addition , for each frequency band considered and at each connection density threshold , the normalised amount of mutual information [44] was calculated between the community structures in the graphs of each pair of subjects . Unlike some previous applications of graph theory to MRI data [33] , [62] , [63] , we did not binarise the thresholded weighted graphs , to be able to better estimate path lengths and between-group differences therein [32] , [64] . However , we verified that all the results described here , except those relating to characteristic path length , remained qualitatively unchanged when calculated with binarised matrices . While the above graph-theoretic measures characterised the topological structure of networks , they did not capture how these networks were embedded in topographical space over scalp . To do this , we calculated a novel measure , termed modular span , which estimated the weighted topographical distance spanned by a module . More formally , given a thresholded graph with a previously identified community structure , the modular span S of a non-degenerate module M ( i . e . , a module with more than one member ) , was defined as:where nM is the number of nodes in the module , and ( i , j ) are a pair of member nodes therein . dij is the normalised Euclidean distance between the pair of corresponding electrodes over the scalp , and wij is the weight of the edge between nodes i and j . Note that , as dij is the normalised distance ( i . e . , dij = 1 for the most distant pair of electrodes ) , modular span is a dimensionless quantity . Modular span as defined above can be interpreted as the weighted sum of the topographic lengths of all the edges between the nodes comprising a module , scaled by the size of the module . By taking an algorithmically derived module of a graph and embedding it in the physical space over the scalp , modular span linked the topological construct with a topographical measure that provided key insights into the spatial differences between the brain networks of patients and controls . We compared the graph metrics described above between groups of patients and controls in frequency bands of interest using unpaired t-tests , assuming unequal variances within the groups . The ability of the metrics derived from individual patient graphs to predict their CRS-R scores was tested using robust linear regression , by calculating R2 and p-values to estimate statistical significance .
|
What are the neural signatures of consciousness ? This is an elusive yet fascinating challenge to current cognitive neuroscience , but it takes on an immediate clinical and societal significance in patients diagnosed as vegetative and minimally conscious . In these patients , it leads us to ask whether we can test for the presence of these signatures in the absence of any external signs of awareness . Recent conceptual advances suggest that consciousness requires a dynamic balance between integrated and differentiated networks of information exchange between brain regions . Here we apply this insight to study such networks in patients and compare them to healthy adults . Using the science of graph theory , we show that the rich and diversely connected networks that support awareness are characteristically impaired in patients , lacking the ability to efficiently integrate information across disparate regions via well-connected hubs . We find that the quality of patients' networks also correlates well with their degree of behavioural responsiveness , and some vegetative patients who show signs of hidden awareness have remarkably well-preserved networks similar to healthy adults . Overall , our research highlights distinctive network signatures of pathological unconsciousness , which could improve clinical assessment and help identify patients who are aware despite being uncommunicative .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"brain",
"electrophysiology",
"electrophysiology",
"neuroscience",
"cognitive",
"neuroscience",
"brain",
"mapping",
"computational",
"neuroscience",
"bioassays",
"and",
"physiological",
"analysis",
"electroencephalography",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"altered",
"states",
"of",
"consciousness",
"clinical",
"neurophysiology",
"consciousness",
"electrophysiological",
"techniques",
"graph",
"theory",
"diagnostic",
"medicine",
"physiology",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"theories",
"of",
"consciousness",
"cognitive",
"science",
"behavioral",
"neuroscience",
"neurophysiology"
] |
2014
|
Spectral Signatures of Reorganised Brain Networks in Disorders of Consciousness
|
Reinforcement Learning has greatly influenced models of conditioning , providing powerful explanations of acquired behaviour and underlying physiological observations . However , in recent autoshaping experiments in rats , variation in the form of Pavlovian conditioned responses ( CRs ) and associated dopamine activity , have questioned the classical hypothesis that phasic dopamine activity corresponds to a reward prediction error-like signal arising from a classical Model-Free system , necessary for Pavlovian conditioning . Over the course of Pavlovian conditioning using food as the unconditioned stimulus ( US ) , some rats ( sign-trackers ) come to approach and engage the conditioned stimulus ( CS ) itself – a lever – more and more avidly , whereas other rats ( goal-trackers ) learn to approach the location of food delivery upon CS presentation . Importantly , although both sign-trackers and goal-trackers learn the CS-US association equally well , only in sign-trackers does phasic dopamine activity show classical reward prediction error-like bursts . Furthermore , neither the acquisition nor the expression of a goal-tracking CR is dopamine-dependent . Here we present a computational model that can account for such individual variations . We show that a combination of a Model-Based system and a revised Model-Free system can account for the development of distinct CRs in rats . Moreover , we show that revising a classical Model-Free system to individually process stimuli by using factored representations can explain why classical dopaminergic patterns may be observed for some rats and not for others depending on the CR they develop . In addition , the model can account for other behavioural and pharmacological results obtained using the same , or similar , autoshaping procedures . Finally , the model makes it possible to draw a set of experimental predictions that may be verified in a modified experimental protocol . We suggest that further investigation of factored representations in computational neuroscience studies may be useful .
Standard Reinforcement Learning ( RL ) [1] is a widely used normative framework for modelling conditioning experiments [2] , [3] . Different RL systems , mainly Model-Based and Model-Free systems , have often been combined to better account for a variety of observations suggesting that multiple valuation processes coexist in the brain [4]–[6] . Model-Based systems employ an explicit model of consequences of actions , making it possible to evaluate situations by forward inference . Such systems best explain goal-directed behaviours and rapid adaptation to novel or changing environments [7]–[9] . In contrast , Model-Free systems do not rely on internal models and directly associate values to actions or states by experience such that higher valued situations are favoured . Such systems best explain habits and persistent behaviours [9]–[11] . Of significant interest , learning in Model-Free systems relies on a computed reinforcement signal , the reward prediction error ( RPE ) . This signal parallels the observed shift of dopamine neurons' response from the time of an initially unexpected reward – an outcome that is better or worse than expected – to the time of the conditioned stimulus that precedes it , which , in Pavlovian conditioning experiments , is fully predictive of the reward [12] , [13] . However recent work by Flagel et al . [14] , raises questions about the exclusive use of classical RL Model-Free methods to account for data in Pavlovian conditioning experiments . Using an autoshaping procedure , a lever-CS was presented for 8 seconds , followed immediately by delivery of a food pellet into an adjacent food magazine . With training , some rats ( sign-trackers; STs ) learned to rapidly approach and engage the lever-CS . However , others ( goal-trackers; GTs ) learned to approach the food magazine upon CS presentation , and made anticipatory head entries into it . Furthermore , in STs , phasic dopamine release in the nucleus accumbens , measured with fast scan cyclic voltammetry , matched RPE signalling , and dopamine was necessary for the acquisition of a sign-tracking CR . In contrast , despite the fact that GTs acquired a Pavlovian conditioned approach response , this was not accompanied with the expected RPE-like dopamine signal , nor was the acquisition of a goal-tracking CR blocked by administration of a dopamine antagonist ( see also [15] ) . Classical dual systems models [16]–[19] should be able to account for these behavioural and pharmacological data , but the physiological data are not consistent with the classical view of RPE-like dopamine bursts . Based on the observation that STs and GTs focus on different stimuli in the environment , we suggest that the differences observed in dopamine recordings may be due to an independent valuation of each stimulus . In classical RL , valuation is usually done at the state level . Stimuli , embedded into states – snapshots of specific configurations in time – , are therefore hidden to systems . In this case , it would prevent dealing separately with the lever and the magazine at the same time . However , such data may still be explained by a dual systems theory , when extended to support and benefit from factored representations; that is , learning the specific value of stimuli independently from the states in which they are presented . In this paper , we present and test a model using a large set of behavioural , physiological and pharmacological data obtained from studies on individual variation in Pavlovian conditioned approach behaviour [14] , [20]–[25] . It combines Model-Free and Model-Based systems that provide the specific components of the observed behaviours [26] . It explains why inactivating dopamine in the core of the nucleus accumbens or in the entire brain results in blocking specific components and not others [14] , [25] . By weighting the contribution of each system , it also accounts for the full spectrum of observed behaviours ranging from one extreme – sign-tracking – to the other [26] – goal-tracking . Above all , by extending classical Model-Free methods with factored representations , it potentially explains why the lever-CS and the food magazine might acquire different motivational values in different individuals , even when they are trained in the same task [22] . It may also account for why the RPE-like dopaminergic responses are observed in STs but not GTs , and also the differential dependence on dopamine [14] .
Not only have Flagel et al . [14] provided behavioural data but they also provide physiological and pharmacological data . This raises the opportunity to challenge the model at different levels , as developed in the current and next sections . Using Fast Scan Cyclic Voltammetry ( FSCV ) in the core of the nucleus accumbens they recorded the mean of phasic dopamine ( DA ) signals upon CS ( lever ) and US ( food ) presentation . It was observed that depending on the subgroup of rats , distinct dopamine release patterns emerge ( see Figure 7 A , B ) during Pavlovian training . STs display the classical propagation of a phasic dopamine burst from the US to the CS over days of training and the acquisition of conditioned responding ( see Figure 7 A ) . This pattern of dopamine activity is similar to that seen in the firing of presumed dopamine cells in monkeys reported by Schultz and colleagues [12] and interpreted as an RPE corresponding to the reinforcement signal of Model-Free RL systems [1] . In GTs , however , a different pattern was observed . Initially there were small responses to both the CS and US , of which the amplitudes seemed to follow a similar trend over training ( see Figure 7 B ) . By recording the mean of the RPEs computed in the Feature-Model-Free system during the autoshaping simulation ( i . e . only fitted to behavioural data ) , the model can still qualitatively reproduce the different patterns observed in dopamine recordings for STs and GTs ( see Figure 7 C , D ) . For STs , the model reproduces the progressive propagation of from the US to the CS ( see Figure 7 C ) . For GTs , it reproduces the absence of such propagation . The RPE at the time of the US remains over training , while a also appears at the time of the CS ( see Figure 7 D ) . In the model , such discrepancy is explained by the difference in the values that STs and GTs use for the computation of RPEs at the time of the CS and the US . STs , by repeatedly focusing on the lever , propagate the total value of food to the lever and end up having a unique at the unexpected lever appearance only . By contrast , by repeatedly focusing on the magazine during the lever appearance but , as all rats , also from time to time during ITI , GTs revise the magazine value multiple times , positively just after food delivery and negatively during ITI . Such revisions lead to a permanent discrepancy between the expected and observed value , i . e . a permanent , at lever appearance and food delivery , when engaging with the magazine . The key mechanism to reproduce these results resides in the generalization capacities of the Feature-Model-Free system . Based on features rather than states , feature-values are to be used , and therefore revised , at different times and states of the experiment , favouring the appearance of RPEs . Variants 2 , 3 and 4 relying on classical Model-Free systems are unable to reproduce such results ( see Figure S3 ) . By using values over abstract states rather than stimuli , it makes it impossible to only revise the value of the magazine during ITI . Therefore , given the deterministic nature of the MDP , we observe a classical propagation of RPEs in all pathways up to the appearance of the lever .
Combining Model-Based and Model-Free systems has previously been successful in explaining the shift from goal-directed to habitual behaviours observed in instrumental conditioning [17]–[19] , [33] , [34] . However , few models based on the same concept have been developed to account for Pavlovian conditioning [16] . While the need for two systems is relevant in instrumental conditioning given the distinct temporal engagement of each system , such a distinction has not been applied to Pavlovian phenomena ( but see recent studies on orbitofrontal cortex [37]–[39] ) . The variability of behaviours and the need for multiple systems have been masked by focusing on whole populations and , for the most part , ignoring individual differences in studies of Pavlovian conditioning . The nature of the CS is especially important , as many studies of Pavlovian conditioned approach behaviour have used an auditory stimulus as the CS , and in such cases only a goal-tracking CR emerges in rats [40] , [41] . As expected from the behavioural data , combining two learning systems was successful in reproducing sign- and goal-tracking behaviours . The Model-Based system , learning the structure of the task , favours systematic approach towards the food magazine , and waiting for food to be delivered , and hence the development of a goal-tracking CR . The Feature-Model-Free system , directly evaluating features by trials and errors , favours systematic approach towards the lever , a full predictor of food delivery , and hence the development of a sign-tracking CR . Moreover , utilizing the Feature-Model-Free system to represent sign-tracking behaviour yields results consistent with the pharmacological data . Disrupting RPEs , which reflects the effects of flupentixol on dopamine , blocks the acquisition of a sign-tracking CR , but not a goal-tracking CR . The model does not make a distinction between simple approach behaviour versus consumption-like engagement , as reported for both STs and GTs [23] , [24] . However given that such engagement results from the development of incentive salience [23] , [24] , the values learned by the Feature-Model-Free system to bias behaviour towards stimuli attributed with motivational value are well-suited to explain such observations . The higher motivational value attributed to the lever by STs relative to GTs can also explain why the lever-CS is a more effective conditioned reinforcer for STs than for GTs [22] . Importantly , none of the systems are dedicated to a specific behaviour , nor rely on a priori information to guide their processes . The underlying mechanisms increasingly make one behaviour more pronounced than the other through learning . Each system contributes to a certain extent to sign- and goal-tracking behaviour . This property is emphasized by the weighted sum integration of the values computed by each system before applying the softmax action-selection mechanism . The variability of behaviours in the population can then be accounted for by adjusting the weighting parameter from ( i . e . favouring sign-tracking ) to ( i . e . favouring goal-tracking ) . This suggests that the rats' actions result from some combination of rational and impulsive processes , with individual variation contributing to the weight of each component . The integration mechanism is directly inspired by the work of Dayan et al . [16] and as the authors suggest , the parameter may fluctuate over time , making the contribution of the two systems vary with experience . In contrast to their model , however , the model presented here does not assign different goals to each system . Thus , the current model is more similar to their previous model [17] , which uses another method for integration . A common alternative to integration when using multiple systems [17] , [18] , [35] is to select at each step , based on a given criterion ( certainty , speed/accuracy trade-off , energy cost ) , a single system to pick the next action . Such switch mechanism does not fit well with the present model , given that it would be interpreted as if actions relied sometimes only on motivational values ( i . e . Feature-Model-Free system ) and sometimes only on a rational analysis of the situation ( i . e . Model-Based system ) . It also does not fit well with pharmacological observation that STs do not express goal-tracking tendencies in the drug-free test session following systemic-injections of flupentixol [14] , as Flagel et al . stated , “[sign-tracking] rats treated with flupentixol did not develop a goal-tracking CR” . Classical RL algorithms used in neuroscience [16]–[18] , [35] , designed mainly to account for instrumental conditioning , work at the state level . Tasks are defined as graphs of states , and corresponding models are unaware of any similarity within states . Therefore , any subsequent valuation process cannot use any underlying structure to generalize updates to states that share stimuli . Revising the valuation process to handle features rather than states per se , makes it possible to attribute motivational values to stimuli independently of the states in which they are presented . Recent models dedicated to Pavlovian conditioning [36] , [42]–[46] usually represent and process stimuli independently and can be said to use factored representations , a useful property to account for phenomena such as blocking [47] or overexpectation [48] . In contrast to the present model , while taking inspiration from RL theory ( e . g . using incremental updates ) , these models are usually far from the classical RL framework . Of significant difference with the present study , most of these models tend to describe the varying intensity of a unique conditioned response and do not account for variations in the actual form of the response , as we do here . In such models , the magazine would not be taken into account and/or taken as part of the context , making it unable to acquire a value for itself nor be the focus of a particular response . In RL theory , factorization is mainly evoked when trying to overcome the curse of dimensionality [49] ( i . e . standard algorithms do not scale well to high dimensional spaces and require too much physical space or computation time ) . Amongst methods that intend to overcome this problem are value function approximations and Factored Reinforcement Learning . Value function approximations [35] , [50] , [51] attempt to split problems into orthogonal subproblems making computations easier and providing valuations that can then be aggregated to estimate the value of states . Factored Reinforcement Learning [52]–[54] attempts to find similarities between states so that they can share values , reducing the physical space needed and relies on factored Markov Decision Processes . We also use factored Markov Decision processes , hence the “factored” terminology . However , our use of factored representations serves a different purpose . We do not intend to build a compact value-function nor infer the value of states from values of features but rather make these values compete in the choice for the next action . Taking advantage of factored representations into classical RL algorithms is at the very heart of the present results . By individually processing stimuli within states ( i . e . in the same context , at the same time and same location ) and making them compete , the Feature-Model-Free system favours a different policy – oriented towards engaging with the most valued stimuli – ( sign-tracking ) than would have been favoured by classical algorithms such as Model-Based or Model-Free systems ( goal-tracking ) . Hence , combining a classical RL algorithm with the Feature-Model-Free system enables the model to reproduce the difference in behaviours observed between STs and GTs during an autoshaping procedure . Moreover , by biasing expected optimal behaviours towards cues with motivational values ( incentive salience ) , it is well suited to explain the observed commitment to unnecessary and possibly counter-productive actions ( see also [16] , [55] , [56] ) . Most of all , it enables the model to replicate the different patterns of dopamine activity recorded with FSCV in the core of the nucleus accumbens of STs and GTs . The independent processing of stimuli leads to patterns of RPE that match those of dopamine activity for STs – a shift of bursts from the US to the CS; and in GTs – a persistence of bursts at both the time of the US and the CS . By combining the two concepts of dual learning systems and factored representations in a single model , we are able to reproduce individual variation in behavioural , physiological and pharmacological effects in rats trained using an autoshaping procedure . Interestingly , our approach does not require a deep revision of mechanisms that are extensively used in our current field of research . While Pavlovian and instrumental conditioning seem entangled in the brain [57] , the two major concepts on which rely their respective models , dual learning systems and factored representations , have to our knowledge never been combined into a single model in this field of research . This approach could contribute to the understanding of interactions between these two classes of learning , such as CRE or Pavlovian-Instrumental Transfer ( PIT ) , where motivation for stimuli acquired via Pavlovian learning modulates the expression of instrumental responses . Interestingly , the Feature-Model-Free system nicely fits with what would be expected from a mechanism contributing to general PIT [58] . It is focused on values over stimuli without regard to their nature [58] , it biases and interferes with some more instrumental processes [55] , [56] , [58] and it is hypothesized to be located in the core of the nucleus accumbens [58] . It would thus be interesting to study whether future simulations of the model could explain and help better formalize these aspects of PIT . We do not necessarily imply that instrumental and Pavlovian conditioning might rely on a unique model . Rather , we propose that if they were the results of separated systems , they should somehow rely on similar representations and valuation mechanisms , given the strength of the observed interactions . The proposed model explains the persistent dopamine response to the US in GTs over days of training as a permanent RPE due to the revision of the magazine value during each ITI . Therefore , a prediction of the model is that shortening the ITI should reduce the amplitude of this burst ( i . e . there should be less time to revise the value and reduce the size of the RPE ) ; whereas increasing the ITI should increase the amplitude of this burst . Removing the food dispenser during ITI , similar to theoretically suppressing the ITI , should make this same burst disappear . Studying physiological data by grouping them given the duration of the preceding ITI might be sufficient , relatively to noise , to confirm that its duration impacts the amplitude of dopamine bursts . In the current experimental procedure , the ITI is indeed randomly picked in a list of values with an average of 90 sec . Moreover , reducing ITI duration should lead to an increase of the tendency to goal-track in the overall population . Indeed , with a higher value of the food magazine , the Feature-Model-Free system would be less likely to favour sign-tracking over goal-tracking CR . The resulting decrease in sign-tracking in the overall population would be consistent with findings of previous works [59]–[62] , where a shorter ITI reduces the observed performance in the acquisition of sign-tracking CRs . Alternatively , it would also be interesting to examine the amplitude of dopamine bursts during the ITI ( especially when exploring the food magazine ) , to determine whether or not physiological responses during this period affect the outcome of the conditioned response . It would be interesting to split physiological data not only between STs and GTs but also between the stimuli on which the rats started and/or ended focusing on during CS presentation at each trial . This would help to confirm that the pattern of dopamine activity is indeed due to a separate valuation of each stimuli . We would predict that at the time of the US , dopamine bursts during engagement with the lever should be small relatively to dopamine bursts during engagement with the magazine . Moreover , comparing dopamine activity at the time of the CS when engaging with the lever versus the magazine could help elucidate which update mechanism is being used . If activity differs , this would suggest that the model should be revised to use SARSA-like updates , i . e . taking into account the next action in RPE computation . Such a question has already been the focus of some studies on dopamine activity [63]–[65] . There is no available experimental data for the phasic dopaminergic activity of the intermediate group . The model predicts that such a group would have a permanent phasic dopamine burst , i . e . RPE , at US and a progressively appearing burst at CS ( see Figure S6 ) . Over training , the amplitude of the phasic dopamine burst at US should decrease until a point of convergence , while at the mean time the response at CS should increase until reaching a level higher than the one observed at US . However , one must note , that the fitting of the intermediate group is not as good as for STs or GTs , as it regroups behaviours that range from sign-tracking to goal-tracking , such that this is a weak prediction . There is the possibility that regularly presenting the magazine or the lever could , without pairing with food , lead to responses that are indistinguishable from CRs . However , ample evidence suggests that the development of a sign-tracking or goal-tracking CR is not due to this pseudoconditioning phenomenon , but rather a result of learned CS-US associations . That is , experience with lever-CS presentations or with food US does not account for the acquisition of lever-CS induced directed responding [22] , [66] . Nonetheless , it should be noted that the current model cannot distinguish between pseudoconditioning CR-like responses and sign-tracking or goal-tracking behaviours . This would require us to introduce more complex MDPs that embed the ITI and can more clearly distinguish between approach and engagement . The Feature-Model-Free system presented in this article was designed as a proof of concept for the use of factored representations in computational neuroscience . In its present form it updates the value of one feature ( the focused one ) at a time , and this is sufficient to account for much of the experimental data . It does not address whether multiple features could be processed in parallel , such that multiple synchronized , but independently computed , signals would update distinct values relative to the attention paid to the associated features . Further experiments should be performed to confirm this hypothesis . Subsequently , using factored representations in the Model-Based system was not necessary to account for the experimental data and the question remains whether explaining some phenomena would require it . While using factored representations , our approach still relies on the discrete-time state paradigm of classical RL , where updates are made at regular intervals . Although such simplification can explain the set of data considered here , one would need to extend this to continuous time if one would like to also model experimental data where rats take more or less time to initiate actions that can vary in duration [14] . The present model , which does not take timing into consideration , cannot account for the fact that STs and GTs both come to approach their preferred stimuli faster and faster as a function of training nor does it make use of the variations of ITI duration . Our attempt to overcome this limitation using the MDP framework was unsuccessful . Focusing on features , it becomes more tempting to deal with the timing of their presence , a property that is known to be learned and to have some impact on behaviours [61] , [67]–[69] . Moreover , in the current model , we did not attempt to account for the conditioned orienting responses ( i . e . orientation towards the CS ) that both STs and GTs exhibit upon CS presentation [25] . However , we hypothesize that such learned orienting responses could be due to state discrimination mechanisms that are not included in the model , and would be better explained with partial observability and actions dedicated to collect information . This is beyond the scope of the current article , but is of interest for future studies . As evident by the only partial reproduction of the flupentixol effects on the expression of sign- and goal-tracking behaviours , the model is limited by the use of the softmax action-selection mechanism , which is widely used in computational neuroscience [16]–[19] , [32] , [34]–[36] . In the model , all actions are equal – there is no action with a specific treatment – and the action-selection mechanism necessarily selects an action at each time step . Any reduction in the value of one action favours the selection of all other actions in proportion to their current associated values . In reality , however , blocking the expression of an action would certainly lead mainly to inactivity rather than necessarily picking the alternative and almost never expressed action . One way of improving the model in this direction could be to replace the classical softmax function by a more realistic model of action selection in the basal ganglia ( e . g . [70] ) . In such a model , no action is performed when no output activity gets above a certain threshold . Humphries et al . [32] have shown that changing the exploration level in a softmax function can be equivalent to changing the level of tonic dopamine in the basal ganglia model of Gurney et al . [70] . Interestingly , in the latter model , reducing the level of tonic dopamine results in difficulty in initiating actions and thus produces lower motor behaviour , as is seen in Parkinsonian patients and as can be seen in rats treated with higher doses of flupentixol [14] . Thus a natural sequel to the current model would be to combine it with a more realistic basal ganglia model for action selection . We simulated the effect of flupentixol as a reduction of the RPE in the learning processes of Model-Free systems to parallel its blockade of the dopamine receptors . While this is sufficient to account for the pharmacological results previously reported [14] , it fails to account for some specific aspects that have more recently emerged . Mainly , it is unable to reproduce the instant decreased engagement observed at the very first trial after post-training local injections of flupentixol [25] . Our current approach requires re-learning to see any impact of flupentixol . A better understanding of the mechanisms that enable instant shifts in motivational values , by shifts in the motivational state [71] or the use of drugs [14] , [25] , might be useful to extend the model on such aspects . We also tried to model the effect of flupentixol on RPEs with a multiplicative effect , as it would have accounted for an instant impact on behaviour . However , it failed to account for the effects of flupentixol on learning of the sign-tracking CRs , as a multiplicative effect only slowed down learning but did not disrupt it . How to model the impact of flupentixol , and dopamine antagonists or drugs such as cocaine remains an open question ( e . g . see [72] , [73] ) . Finally , our work does not currently address the anatomical counterpart of at the heart of the model , nor the regions of the brain that would match the current Model-Based system and the Feature-Model-Free system . Numerous studies have already discussed the potential substrates of Model-Based/Model-Free systems in the prefrontal cortex/dorsolateral striatum [74] , or the dorsomedial and dorsolateral striatum [33] , [75]–[78] . The weighted sum integration may suggest a crossed projection of brains regions favouring sign- and goal-tracking behaviours ( Model-Based and Feature-Model-Free systems ) into a third one . We postulate there is a difference in strength of “connectivity” between such regions in STs vs GTs [79] . Further , one might hypothesize that the core of the nucleus accumbens contributes to the Feature-Model-Free system . The integration and action selection mechanisms would naturally fit within the basal ganglia , stated to contribute to such functions [32] , [80]–[82] . Here we have presented a model that accounts for variations in the form of Pavlovian conditioned approach behaviour seen during autoshaping in rats; that is , the development of a sign-tracking vs goal-tracking CR . This works adds to an emerging set of studies suggesting the presence and collaboration of multiple RL systems in the brain . It questions the classical paradigm of state representation and suggests that further investigation of factored representations in RL models of Pavlovian and instrumental conditioning experiments may be useful .
In the classical reinforcement learning theory [1] , tasks are usually described as Markov Decision Processes ( MDPs ) . As the proposed model is based on RL algorithms , we use the MDP formalism to computationally describe the Pavlovian autoshaping procedure used in all simulations . An MDP describes the interactions of an agent with its environment and the rewards it might receive . An agent being in a state can execute an action which results in a new state and the possible retrieval of some reward . More precisely , an agent can be in a finite set of states , in which it can perform a finite set of discrete actions , the consequences of which are defined by a transition function , where is the probability distribution of reaching state doing action in state . Additionally , the reward function is the reward for doing action in state . Importantly , MDPs should theoretically comply with the Markov property: the probability of reaching state should only depend on the last state and the last action . An MDP is defined as episodic if it includes at least one state which terminates the current episode . Figure 1 shows the deterministic MDP used to simulate the autoshaping procedure . Given the variable time schedule ( 30–150s ) and the net difference observed in behaviours in inter-trial intervals , we can reasonably assume that each experimental trial can be simulated with a finite horizon episode . The agent starts from an empty state ( ) where there is nothing to do but explore . At some point the lever appears ( ) and the agent must make a critical choice: It can either go to the lever ( ) and engage with it ( ) , go to the magazine ( ) and engage with it ( ) or just keep exploring ( , ) . At some point , the lever is retracted and food is delivered . If the agent is far from the magazine ( , ) , it first needs to get closer . Once close ( ) , it consumes the food . It ends in an empty state ( ) which symbolizes the start of the inter-trial interval ( ITI ) : no food , no lever and an empty but still present magazine . The MDP in Figure 1 is common to all of the simulations and independent of the reinforcement learning systems we use . STs should favour the red path , while GTs should favour the shorter blue path . All of the results rely mainly on the action taken at the lever appearance ( ) , when choosing to go to either the lever , the magazine , or to explore . Exploring can be understood as not going to the lever nor to the magazine . To fit with the requirements of the MDP framework , we introduce two limitations in our description , which also simplify our analyses . We assume that engagement is necessarily exclusive to one or no stimulus , and we make no use of the precise timing of the procedure – the ITI duration nor the CS duration – in our simulations . The model relies on the architecture shown in Figure 2 . The main idea is to combine the computations of two distinct reinforcement learning systems to define what behavioural response is chosen at each step . Given the modular architecture of the model , we were able to test different combinations of RL systems . Their analysis underlined the key mechanisms required for reproducing each result ( see Figures S1 , S2 , S4 and S5 ) . Figure 11 ( B , C and D ) schematically represents the analysed variants . Most of the results rely on the action taken by the agent at the lever appearance . The action taken results from the values , and , the computation of which differs in each of the variants described below . The model relies on model-specific parameters ( , , and ) and experience-specific parameters ( , , and ) . If the model were used to simulate a different experiment , the model-specific parameters would be the same while different experience-specific parameters might be required . For an easier analysis and a simpler comparison between the model and its variants , we reduce the number of parameters by sharing parameters with identical meanings amongst systems ( i . e . both systems within the model share values for their learning rates and discount rates , rather than having independent parameter values ) . Due to the number of parameters , finding the best values to qualitatively fit the experimental data cannot be done by hand . Using a genetic algorithm makes it possible to optimize the search of suitable values for the parameters . Parameter values were retrieved by fitting the simulation of the probabilities to engage either the lever or the magazine with the experimental data of one of the previous studies [21] . No direct fitting was intended on other experimental data . Hence , a single set of values was used to simulate behavioural , physiological and pharmacological data . If for a variant , the optimization algorithm fails to fit the experimental data , it suggests that whatever the values , the mechanisms involved cannot explain the behavioural data ( Variant 4 ) . Probabilities to engage the lever or the magazine were taken as independent objectives of the algorithm , since fitting sign-tracking probabilities is easier than fitting goal-tracking probabilities . For each objective , the fitness function is computed as the least square errors between the experimental and simulated data . Parameter optimization is done with the multi-objective genetic algorithm NSGA-II [86] . We used the implementation provided by the Sferes 2 framework [87] . All parameters required for reproducing the behavioural data were fitted at once . For NSGA-II , we arbitrarily use a population of 200 individuals and run it over 1000 generations . We use a polynomial mutation with a rate of 0 . 1 , and simulate binary cross-overs with a rate of 0 . 5 . We select the representative individual , to be displayed in figures , from the resulting Pareto front by hand , such that it best visually fits the observed data . To confirm that is the key parameter of the model , we additionally tried to fit the whole population at once ( i . e . sharing all parameter values in agents but ) and we were still able to reproduce the observed tendencies of sign- and goal-tracking in the population ( see Figure S7 A , B ) and the resulting different phasic dopaminergic patterns ( see Figure S7 C , D ) . It is however almost certain that each subgroup does not express the exact same values for the other parameters . Removing such constraint by fitting each subgroup separately , indeed provides better results . Results presented in this article are based on such separate fitting .
|
Acquisition of responses towards full predictors of rewards , namely Pavlovian conditioning , has long been explained using the reinforcement learning theory . This theory formalizes learning processes that , by attributing values to situations and actions , makes it possible to direct behaviours towards rewarding objectives . Interestingly , the implied mechanisms rely on a reinforcement signal that parallels the activity of dopamine neurons in such experiments . However , recent studies challenged the classical view of explaining Pavlovian conditioning with a single process . When presented with a lever whose retraction preceded the delivery of food , some rats started to chew and bite the food magazine whereas others chew and bite the lever , even if no interactions were necessary to get the food . These differences were also visible in brain activity and when tested with drugs , suggesting the coexistence of multiple systems . We present a computational model that extends the classical theory to account for these data . Interestingly , we can draw predictions from this model that may be experimentally verified . Inspired by mechanisms used to model instrumental behaviours , where actions are required to get rewards , and advanced Pavlovian behaviours ( such as overexpectation , negative patterning ) , it offers an entry point to start modelling the strong interactions observed between them .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"behavioral",
"neuroscience",
"computational",
"neuroscience",
"biology",
"neuroscience",
"learning",
"and",
"memory"
] |
2014
|
Modelling Individual Differences in the Form of Pavlovian Conditioned Approach Responses: A Dual Learning Systems Approach with Factored Representations
|
In most organisms , the heat-shock response involves increased heat-shock gene transcription . In Kinetoplastid protists , however , virtually all control of gene expression is post-transcriptional . Correspondingly , Trypanosoma brucei heat-shock protein 70 ( HSP70 ) synthesis after heat shock depends on regulation of HSP70 mRNA turnover . We here show that the T . brucei CCCH zinc finger protein ZC3H11 is a post-transcriptional regulator of trypanosome chaperone mRNAs . ZC3H11 is essential in bloodstream-form trypanosomes and for recovery of insect-form trypanosomes from heat shock . ZC3H11 binds to mRNAs encoding heat-shock protein homologues , with clear specificity for the subset of trypanosome chaperones that is required for protein refolding . In procyclic forms , ZC3H11 was required for stabilisation of target chaperone-encoding mRNAs after heat shock , and the HSP70 mRNA was also decreased upon ZC3H11 depletion in bloodstream forms . Many mRNAs bound to ZC3H11 have a consensus AUU repeat motif in the 3′-untranslated region . ZC3H11 bound preferentially to AUU repeats in vitro , and ZC3H11 regulation of HSP70 mRNA in bloodstream forms depended on its AUU repeat region . Tethering of ZC3H11 to a reporter mRNA increased reporter expression , showing that it is capable of actively stabilizing an mRNA . These results show that expression of trypanosome heat-shock genes is controlled by a specific RNA-protein interaction . They also show that heat-shock-induced chaperone expression in procyclic trypanosome enhances parasite survival at elevated temperatures .
When living organisms are exposed to temperatures above their growth optima , they respond by increased synthesis of heat-shock proteins . In eukaryotes as diverse as animals , ciliates and plants , heat-shock protein expression is controlled by heat-shock transcription factors , whose activation enables them to bind conserved heat-shock elements in the promoters of heat-shock protein genes and activate their transcription [1] , [2] , [3] , [4] . Trypanosoma brucei and related Kinetoplastid protists must also adapt to different temperatures: they multiply both in mammals , with temperatures varying from 32°C to 38°C depending on species and body location ( see e . g . [5] , [6] , [7] ) , and in arthropod vectors in which the temperature variations are much greater ( e . g . [8] ) . In Kinetoplastids , however , the regulation relies exclusively on post-transcriptional mechanisms . Transcription is polycistronic [9] , [10] , and individual mRNAs are produced by trans splicing and polyadenylation [11] , [12] . The final cytoplasmic RNA level is determined by the rates of processing , transport from the nucleus , and degradation [13] . For most trypanosome mRNAs , the rate of degradation is a critical determinant of expression [14] . Two forms of T . brucei are routinely studied in the laboratory: the bloodstream form ( found in the mammalian host , cultivated axenically in vitro at 37°C ) and the procyclic form ( found in the midgut of the Tsetse fly vector , cultivated axenically at 27°C ) . Upon transfer of procyclic forms to 41°C , transcription by RNA polymerase II is gradually shut down [15] and trans splicing is inhibited [16] . The overall level of translation also decreases , as shown by reduced in vivo [35S]-methionine labelling and the collapse of polysome profiles [17] . This is partly due to rapid mRNA degradation , as judged both by profiling of total mRNA [17] and examination of specific transcripts [18]; and it is partly due to effects on translation [17] . After heat-shock , poly ( A ) binding protein and several translation factors accumulate in granules [17] . Meanwhile , the mRNAs encoding HSP83 and the major cytosolic HSP70 remain stable and continue to be translated [17] , [18] . Although the trypanosomes are able to recover from a 41°C heat shock lasting up to 2 hours , it is not known whether the heat-shock response is required for the recovery . Indeed , it is not known whether the trypanosome heat-shock response has any selective advantage . T . brucei has five virtually identical genes encoding the major cytosolic HSP70 that are arranged in a tandem array [19] ( unfortunately collapsed to one locus , Tb927 . 11 . 11330 , in the genome assembly ) and are constitutively co-transcribed [20] , [21] . Using reporters , it was shown that sequence elements in the HSP70 3′-untranslated region ( 3′-UTR ) are responsible for the stability of the mRNA after heat-shock [18] , [22] . Similar observations were also made for HSP70s of the Kinetoplastids Trypanosoma cruzi [23] and Leishmania infantum [24] , [25] . The multiple copies of the HSP83 genes ( encoding the major Kinetoplastid HSP90 homologue ) are also in a tandem array . The 3′-UTR of Leishmania HSP83 mRNA is important for both mRNA stability and increased translation during heat-shock [26] , and it was proposed that temperature-induced changes in RNA secondary structure might play a role in regulation [27] . Post-transcriptional mechanisms are also responsible for heat-induced increases in Leishmania HSP100 mRNA [28] , [29] . The stability , localization and translation states of eukaryotic mRNAs are influenced by proteins that bind to them . For example , in mammalian cells , tristetraprolin ( also called TTP , Tis11a , and Zfp36 ) , and BRF1 and BRF2 ( Butyrate response factors 1 and 2 ) bind to AU-rich elements with a consensus of UAUUUAUU; they recruit components of the mRNA degradation machinery , promoting mRNA decay [30] . These three proteins , together with related proteins from other Opisthokonts ( together called the “Tis11 family” ) , possess two C8C5C3H zinc finger domains separated by a linker of about 10 amino acids . Immediately preceding the zinc finger domain is a six-residue conserved sequence , R/K-Y-K/R-T-E/K-L , which strongly influences the sequence specificity of RNA binding [31] . The activities of TTP and BRF proteins are regulated by phosphorylation , and are critical for control of inflammation and cell proliferation in mammals [30] . Other proteins compete for binding to the AU-rich element and promote mRNA stability [32] . T . brucei has forty-nine CCCH zinc finger proteins , some of which have been implicated in control of gene expression [33] , [34] , [35] , [36] , [37] , [38] , [39] . So far , however , none has been shown to have a destabilising function . In search of possible destabilising proteins , we looked for predicted trypanosome proteins with the Tis11 consensus . We here show that the protein with the best match , ZC3H11 , indeed binds to mRNAs containing an AUU sequence element but that - in contrast to the situation in mammalian cells - the consequence is an increase in mRNA abundance . Most interestingly , ZC3H11 appears to be a master regulator of stress response mRNAs .
To find CCCH proteins that might be involved in post-transcriptional gene regulation in T . brucei , we scanned all of them for the Tis11 consensus . The best matches were in ZC3H11 , ZC3H12 and ZC3H13; ZFPs 1–3 also showed some similarity ( Figure 1A ) . ZC3H11 ( locus Tb927 . 5 . 810 ) consists of 364 amino acids , and has a predicted molecular weight of 39 . 6 kDa . The zinc finger starts at residue 70 , and is preceded by the Tis11 consensus RYKTKL . The ZC3H11 gene is found in all available Kinetoplastid genomes . Most sequence identity is concentrated around the zinc finger: comparing all available proteomes , the 6mer has consensus RYKTK ( L/Y/F ) . Some additional conserved patches are the sequence H ( N/D ) PY around residue 200 of T . brucei ZC3H11 and a serine-rich region near the C-terminus ( Supplementary Figure S1 ) . Alignment of the ZC3H11 zinc finger with those of other Tis11 family proteins showed that some of the residues required for interaction with AU-rich elements were conserved . From a crystal structure of BRF2 ( Tis11-d ) with UUAUUUAUU [31] , it was found that each zinc finger of BRF2 specifically binds the sequence UAUU . Using the residue numbering in Figure 1A , and numbering the 4-nt bound RNA as U ( 1 ) -A ( 2 ) -U ( 3 ) -U ( 4 ) , Tyr18 intercalates between the Us in the 3rd and 4th positions ( U3 and U4 ) , and Phe26 intercalates between U1 and A2 . These residues are conserved in ZC3H11 . Specificity for U1 was conferred by backbone hydrogen bonds with ( Asn/His ) 25 and Glu5; these residues are not conserved in ZC3H11 , which has a basic residue at position 5 , like C . elegans MEX5 ( Figure 1A ) . In contrast , A2 is hydrogen-bonded by Leu6 and Arg8 , which are conserved . A notable difference between ZC3H11 and other Tis11-family zinc fingers is the presence of a novel Asp residue at position 21 in place of the conserved glycine . Attempts to generate a polyclonal antibody that could detect ZC3H11 in parasite lysates failed . In order to detect ZC3H11 in trypanosomes , we therefore integrated a sequence encoding an N-terminal V5-epitope tag [40] in frame with one of the ZC3H11 open reading frames ( ORFs ) . Since the 3′-UTR is conserved by this procedure , expression levels are expected to be approximately normal unless the tag affects protein stability . In both procyclic forms , which we normally grow at 27°C , and the bloodstream stage , grown at 37°C , the V5-ZC3H11 fusion protein was detected as an extremely faint band that migrated at about 60 kDa instead of the expected 40 kDa ( Figure 1B & C , lane 2 ) . The abundance of V5-ZC3H11 was , however , dramatically increased upon heat shock . In procyclic forms , induction of ZC3H11 was transient at 37°C ( Figure 1B lanes 3–6 ) , but stronger and more extended at 41°C; at later time points , some smaller products appeared ( Figure 1B lanes 7–10 ) . Since the longer incubations resulted in a decline in cell viability , the faster-migrating bands could indicate either proteolytic degradation or the removal of posttranslational modifications . For bloodstream forms , incubation at 43°C led to a rapid induction of V5-ZC3H11 although the cells started to die within an hour ( Figure 1C lanes 3 , 4 ) . Further experiments showed that in addition to elevated temperatures , mild translational stress from low concentrations of puromycin also increased V5-ZC3H11 expression in both developmental stages ( Figure 1B lane 11–13 and 1C lanes 5 , 6 ) . In a preliminary attempt to determine the mechanism of this expression regulation , we incubated cells with lactacystin ( not shown ) or MG132 ( Figure 1D ) to inhibit the proteasome . Indeed , the amount of ZC3H11 increased ( Figure 1D ) . This might mean that the protein is normally rapidly degraded by the proteasome , but is stabilised upon heat shock or puromycin stress . Alternatively , proteasome inhibition could be acting as another sort of stress , with ZC3H11 protein increasing by another mechanism . We also expressed ZC3H11-myc in cells with V5-ZC3H11 . Reciprocal pull-downs revealed no evidence for dimerization ( not shown ) . To determine the nature of the possible post-translational modifications , we incubated cell lysates with λ-phosphatase before electrophoresis . The 60 kDa bands ( Figure 1E , upper panel , lane 1 ) collapsed to one band at approximately 50 kDa ( Figure 1E , upper panel , lane 3 ) ; this was prevented by addition of phosphatase inhibitors ( Figure 1E , upper panel , lane 4 ) , showing that the 10 kDa migration difference was indeed due to phosphorylation . Interestingly , the extra band seen after heat shock at 41°C had a similar migration ( Figure 1B lanes 8–10 , Figure 1C , lane 3 ) , suggesting that it too might have been dephosphorylated . For an N-terminal fragment containing the first 128 residues of ZC3H11 , extending 39 residues beyond the zinc finger , a similar pattern was observed ( Figure 1E , lower panel ) , except that a portion appeared unmodified . The N-terminal fragment ran as several bands between 23 kDa and 17 kDa , which all collapsed to the lowest band upon phosphatase treatment . This indicates that residues in the N-terminal region can be phosphorylated . The low abundance of the V5-in situ tagged protein precluded localisation by microscopy or cell fractionation . We therefore instead looked at the location of ZC3H11 bearing a tandem affinity purification ( TAP ) tag , inducibly expressed from a strong RNA polymerase I promoter in procyclic forms . ZC3H11-TAP was clearly excluded from the nucleus and found in the cytoplasm in somewhat granular structures ( Figure 1F ) . The presence of an IgG-binding domain in the tag prevented us from looking for colocalisation with stress granule markers . Inducibly expressed ZC3H11 with a C -terminal myc tag gave similar results but with a much fainter signal: a rather granular cytoplasmic immunofluorescence which became marginally brighter after a one-hour heat shock ( Supplementary Figure S1B ) . To find out which mRNAs were bound by ZC3H11 , we inducibly expressed myc-tagged ZC3H11 in procyclic trypanosomes , precipitated the protein using anti-myc antibody , and compared bound and unbound RNAs by RNASeq . The twenty-four most strongly enriched transcripts are listed in Table 1 and the full list is in Supplementary Table S1 , sheet 1 . Strikingly , more than half of the strongly bound mRNAs were implicated in the stress response . Thirteen of them encoded a full set of chaperones required for protein refolding . All classes of cytosolic HSPs were represented - HSP70 , HSP83 ( HSP90 family ) , HSP100 , HSP110 and HSP20 . Also , mRNAs encoding putative homologues of co-chaperones were present: DnaJ ( HSP40 ) proteins , a FKBP/TPR domain protein , stress induced protein 1 ( STI1 ) and cyclophilin-40 . Three additional bound mRNAs encoded the mitochondrial chaperone HSP60 , a copper chaperone for cytochrome c , and the glutaredoxin GRX2 , which protects against oxidative stress . Notably , mRNAs encoding chaperones for co-translational folding ( TRiC complex ) or organellar import ( mitochondrial HSP70 and ER-resident BiP were not enriched . Among bound transcripts that do not encode annotated chaperones , the most notable encoded GPEET procyclin . Five bound mRNAs encoded proteins of unknown function . We next analysed the 3′-UTRs of all bound transcripts for enriched motifs . We found a striking enrichment of an ( AUU ) n repeat motif ( Figure 2A , supplementary Table S1 , sheet 2 ) . Of the 22 most strongly bound mRNAs , 14 contained perfect ( AUU ) 4 repeats and four more had a repeat of 11 nt ( Table 1 ) : they included all but one ( CYP40 ) of the chaperone mRNAs . A scan of the whole genome revealed 325 genes with a good match to the 12mer ( AUU ) 4 sequence in their predicted 3′-UTR , of which only 44 were at least two-fold enriched in the ZC3H11-bound fraction . Although the 3′-UTRs used in the analysis may , in some cases , be incorrect , this result shows that the presence of an ( AUU ) repeat alone is not sufficient to give ZC3H11 binding . The putative AUU repeat binding motif is interesting because the sequence bound by a single Tis-11 CCCH domain is UAUU [31] . To investigate the RNA-binding specificity of ZC3H11 in more detail , we expressed a variety of different fusion proteins in E . coli and purified them . The only proteins that could be obtained in reasonable quantity and purity were two N-terminal fragments of 104 and 119 residues ( Figure 2B , lanes 1 & 2 ) . Both contain the zinc finger but the 119mer also includes additional conserved residues ( Supplementary Figure S1A ) . Although the proteins formed single bands on denaturing gels ( Supplementary Figure S2A ) , on native gels , the pattern was very smeared and some protein remained in the well ( Supplementary Figure S2B ) . This suggested that despite initial solubility , the proteins were not fully folded and some aggregation was occurring . As controls , we expressed the same protein fragments with a C70S mutation in the zinc finger . These were very poorly expressed and as a consequence , the purified samples were heavily contaminated ( Figure 2B , lanes 3 and 4 ) . To test for RNA binding , we incubated the proteins with various radioactively-labelled oligo-ribonucleotides and examined migration in non-denaturing polyacrylamide gels . The 104 residue protein ( ZC3H11-104 ) , interacted with both ( UAUU ) 5UAU ( classical ARE ) and U ( UAU ) 7U to give a clear band ( Figure 2C , lanes 4 & 5 , arrow s1 ) . U ( UCU ) 7U gave a very faint band of slower mobility ( Figure 2C , lane 6 , arrow n ) which was also detected for both U ( UAU ) 7U and U ( UCU ) 7U using the zinc finger mutants ( Figure 2C , lanes 11 , 12 , 14 , 15 ) . This band most likely represents binding of the probe by an E . coli contaminant , although zinc-finger-independent binding by ZC3H11 is also possible . Using the 119-residue protein , ZC3H11-119 , a slower-mobility band was obtained using ( UAUU ) 5UAU and U ( UAU ) 7U ( Figure 2C , lanes 7 & 8 , band s2 ) , suggesting binding of additional copies of the protein: perhaps the extra 15 amino acids mediate protein-protein interactions . In addition , there was strong accumulation of radioactivity in the well ( w ) . This could be explained if the zinc finger were properly folded , but the remainder of the polypeptide were unfolded and formed aggregates . Reducing the probe length to 14 residues did not affect the apparent aggregation ( not shown ) . To investigate the binding in more detail , we used more probes . Results for ZC3H11-104 are shown in Figure 2D and 2E . The interactions with U ( UAU ) 7U and ( UAUU ) 5UAU were confirmed ( Figure 2D , lanes 2 & 4 , arrow s1 ) , and the faint band using U ( UCU ) 7U ( Figure 2D , lane 6 , arrow n ) was much stronger using U23 ( Figure 2D , lane 8 , arrow n ) . A very faint shift was seen with A23 , but none with C23 ( Figure 2D , lanes 10 & 12 ) . To further assess specificity , ZC3H11-104 was incubated with the labelled U ( UAU ) 7U probe in the presence of unlabelled competitors . U ( UAU ) 7U competed effectively , most of the probe now remaining unbound ( Figure 2E , lanes 3 and 4 ) . In contrast , addition of cold ( UAUU ) 5UAU ( Figure 2E , lanes 5 and 6 ) or of U23 shifted the radioactivity to the non-specific band ( Figure 2E , lanes 9 and 10 , band n ) . U ( UCU ) 7U showed partial competition ( Figure 2E , lanes 7 and 8 ) whole A23 and C23 could not compete at all . Results for ZC3H11-119 were similar except that as before , the specific complex showed slower migration and radioactivity accumulated in the well ( Supplementary Figure S2C–E ) . The strong shift to the apparently less specific band ( “n” ) in the competition assays was not inhibited by heparin ( Supplementary Figure S2E ) . We also attempted to assess the binding affinities of ZC3H11-119 to limiting amounts of U ( UAU ) 7U , ( UAUU ) 5UAU and U23 probes . The only probe that bound at all under these conditions was U ( UAU ) 7U , but the results could not be interpreted quantitatively because at most protein concentrations , the only bound radioactivity was stuck in the well ( Supplementary Figure S2F ) . We concluded that the zinc finger of ZC3H11 binds preferentially to ( UAU ) repeats , but is also able to bind to the classical ARE . We examined the effect of ZC3H11 depletion by RNA interference ( RNAi ) . Stable cell lines inducibly expressing a double stranded RNAi fragment were created in bloodstream- and procyclic-form trypanosomes . In bloodstream forms , depletion of ZC3H11 was lethal ( Figure 3A ) , while no effect was observed in procyclic cells ( Figure 3B ) . A similar result was obtained in a published high-throughput RNAi screen [41] . To further investigate the reason why ZC3H11 was essential in bloodstream-form trypanosomes , the transcriptome of ZC3H11-depleted cells was compared with that of wild-type cells , initially using an oligonucleotide microarray ( not shown ) and later , using poly ( A ) + RNA , by high-throughput cDNA sequencing ( RNA-Seq ) ( Supplementary Table S1 , sheet 3 and Supplementary Figure S3A , B ) . The RNA from ZC3H11-depleted cells was taken 24 h after induction of RNAi , before a growth defect was evident , and with no drug treatment apart from tetracycline , which is known not to affect the transcriptome at the level used [14] . We compared the RNASeq results with a previous dataset for poly ( A ) + RNA from wild-type cells . 452 transcripts were at least 2-fold increased after ZC3H11 depletion ( Supplementary Table S1 , sheet 3 ) . The increased transcripts were significantly enriched in the categories of protein kinases and phosphatases , and also RNA-binding proteins , but have not yet been examined further . There was no correlation between the effects of ZC3H11 RNAi in bloodstream forms and enrichment in the ZC3H11-bound fraction , suggesting that many of the effects seen were secondary . Bound RNAs that increased included GPEET procyclin , but since procyclin-associated mRNAs , which are in the same transcription unit , also increased , an increase in procyclin locus transcription is possible . Other increased ZC3H11-bound mRNAs included those encoding the putative RNA-binding protein RBP5 and a few proteins of unknown function . RNASeq revealed 72 genes with at least 2-fold decreased mRNA expression after ZC3H11 RNAi . The strong enrichment for genes encoding ribosomal proteins ( P = 5×10−15 ) and translation factors ( P = 0 . 05 ) suggests that some of the decreases could be indirect effects , secondary to the onset of growth arrest . Looking at ZC3H11-bound mRNAs ( Table 1 ) , one encoding an FKBP-like petidyl-prolyl cis-trans isomerase was not decreased according to the RNASeq , but was decreased by Northern blotting ( 0 . 2× , Supplementary Figure 3C ) and microarray ( 0 . 4× , not shown ) . HSP70 mRNA levels were reproducibly halved by RNASeq , microarray ( not shown ) and Northern blotting ( Figure 4A and Supplementary Figure 4B ) . We therefore decided to investigate HSP70 regulation by ZC3H11 . To confirm binding of HSP70 mRNA to ZC3H11 , we immunoprecipitated ZC3H11-myc from procyclic trypanosome extracts and subjected the resulting RNA to Northern blotting ( Figure 4B ) . As controls , we used cells that expressed no myc-tagged protein , or cells expressing a myc-tagged version of ZC3H11 with the C70S mutation . Since the immunoprecipitation is a lengthy procedure , some degradation of the mRNA occurred , but nevertheless , a band of HSP70 mRNA was visible in the preparation from cells expressing ZC3H11-myc , whereas no HSP70 mRNA was detected in the control pull-downs . As a further control for non-specific RNA sticking to the beads we looked for the highly abundant tubulin mRNA . As expected , some of this mRNA was found in all lanes , but with no specificity for pull-down by ZC3H11-myc ( Figure 4B ) . The C70S mutant protein was rather poorly expressed relative to the wild-type ( not shown ) , so this experiment by itself allows no conclusions regarding a requirement for the C70 residue of the zinc finger in RNA binding . To find out whether the effect on HSP70 mRNA abundance in bloodstream-form trypanosomes was caused by increased instability , we inhibited transcription and measured the amount of HSP70 mRNA left after 15 and 30 min . In five independent measurements , the HSP70 mRNA half life was 23±7 min ( mean ± standard deviation ) . The RNAi cell line yielded values of 21±9 min in the absence of tetracycline , and 15±6 min after one day of RNAi induction . ZC3H11 RNAi decreased the half-life of HSP70 mRNA in every experiment , suggesting that ZC3H11 stabilises HSP70 mRNA . It was already known that trypanosome HSP70 mRNA abundance is regulated by the 3′-UTR [17] , [18] , [22] . To define the region that was targeted by ZC3H11 , we generated bloodstream-form cell lines that had inducible RNAi against ZC3H11 , and also constitutively expressed chloramphenicol acetyltransferase ( CAT ) reporter constructs flanked by different UTRs ( Figure 4C , 4D and Supplementary Figure S4A ) . The constructs were integrated into the tubulin locus and expressed by read-through transcription by RNA polymerase II . Reporter protein expression was measured by the CAT assay; CAT RNA levels and correct mRNA processing were assessed by Northern blotting ( Figure 4D and Supplementary Figure S4 , B & C ) . The parental construct expressed an mRNA with the 5′-UTR from the EP mRNA , and a truncated actin 3′-UTR ( Figure 4D , control ) . Introducing the HSP70 5′-UTR caused no significant change in expression levels compared to the parental constructs ( Figure 4D , HSP70 5′-UTR ) . In contrast , when the HSP70 3′-UTR was included , either by itself or together with the HSP70 5′-UTR , the steady state levels of CAT mRNA and protein were approximately twice the control . Induction of RNAi against ZC3H11 reduced this expression to the level of the control construct . These results show that the HSP70 3′-UTR was sufficient for ZC3H11-mediated regulation . Since all constructs were transcribed from the same locus , the mechanism must be post-transcriptional . We attempted to compare the half-lives of the CAT-HSP70 reporter mRNAs but the low amounts present after ZC3H11 RNAi prevented accurate quantitation . We wanted to see whether the AU-rich sequence was required for ZC3H11-mediated mRNA stabilisation . A reporter with just the 5′ part of the HSP70 3′-UTR , which lacks the AU sequence element , was expressed at levels similar to the control , and showed no response to ZC3H11 RNAi ( Figure 4D , delAU ) . In contrast , a construct containing only the 3′ part , with mainly just the AU element , behaved like the construct with the complete 3′-UTR . We concluded that the part of HSP70 3′-UTR that contains ( AUU ) repeats is necessary and sufficient for regulation by ZC3H11 in bloodstream forms . Finally , we inserted ( TAT ) 6 either at the beginning , or at the end , of the actin 3′-UTR in the reporter plasmid . The insertion after the coding region and before the actin 3′-UTR had no effect on CAT RNA or protein ( Supplementary Figure S4D ) . An insertion just before the usual poly ( A ) site resulted in a two-fold increase in both RNA and protein , but the effect was independent of the 18mer orientation and was not affected by ZC3H11 RNAi ( Supplementary Figure S4D ) . This confirms our impression that in order to respond to ZC3H11 , the AU-rich sequence requires a particular context . Other regulatory elements behave similarly in trypanosomes: for example , EP mRNA degradation in bloodstream-form trypanosomes is regulated by a 26mer [42] , [43] , but the 26mer alone does not work if placed at the start of the actin 3′-UTR [42] . Since chaperones were strongly enriched among the possible ZC3H11 targets , we investigated ZC3H11 function in procyclic trypanosomes incubated above their normal culture temperature of 27°C . At 37°C , wild-type cells stopped multiplying after 3–4 days , while cells with RNAi showed a slower cell number increase and were already starting to die after 2–3 days ( Figure 5A ) . Since the latter result suggested that ZC3H11 was important in survival at elevated temperatures , we tested published heat-shock conditions . After a one-hour incubation of our normal procyclic forms at 41°C , motility was strongly reduced , but , as previously observed [17] , the cells recovered rapidly after being returned to 27°C ( Figure 5B , WT ) . If the cells were depleted of ZC3H11 , however , recovery was severely impaired ( Figure 5B , RNAi ) . In another cell line , containing only one , V5-tagged copy of ZC3H11 , recovery kinetics were intermediate between the RNAi and wild-type ( not shown ) . This suggests that V5-ZC3H11 is functional , but the presence of only a single copy of ZC3H11 causes haplo-insufficiency . To look at the effect of the heat shock on cell cycle progression in more detail , we analysed cell shape and DNA content by FACS ( Figure 5C ) . Normal cells before shock had identical patterns with a G1 peak of 1× diploid DNA content , a smaller G2/M peak with 2× diploid DNA content , and cells in S-phase in between . One day after the heat shock , both populations showed relatively more G2/M cells , an accumulation of multinucleate cells with abnormally high DNA content , and some dead cells with less than 1× DNA content . The wild-type population had returned to normal by day 2 , but for the population with ZC3H11 RNAi , dead cells and cells with abnormally high DNA content persisted and the G1/G2 ratio had not recovered ( Figure 5C ) . As previously described [17] , a one-hour 41°C heat shock reproducibly decreased de novo synthesis of many proteins , as judged by [35S]-methionine labelling ( Figure 5D ) ; among those spared were two migrating at about 90 kDa and 70 kDa , which are probably HSP83 and HSP70 . This result was extremely similar to that previously seen for insect-stage Leishmania [44] , [45] . Transcription initiation is shut down in trypanosomes after heat shock [15] and by preparing RNA , then analysing the amount of mRNA by Northern blotting with a spliced leader probe , we found that the global mRNA level was decreased after the 1 h-heat shock whether or not ZC3H11 RNAi had been induced ( Figure 5E ) . The mRNA encoding alpha tubulin decreased by 20–30% and mRNA encoding glycerol-3-phosphate dehydrogenase by 50% ( Supplementary Figure S5 ) . As expected , in heat-shocked cells without RNAi HSP70 mRNA persisted ( Figure 5F , lanes 1 & 2 ) . In contrast , after ZC3H11 RNAi , stabilisation of HSP70 mRNA was no longer seen ( Figure 5F , lanes 5 & 6 ) . Similar results were observed for HSP83 , HSP110 , FKBP , and the mRNA encoding the HSP40/DnaJ-like protein J2; moreover , HSP100 mRNA was induced by heat shock in wild-type cells but not induced after ZC3H11 RNAi ( Supplementary Figure S5 ) . Cultivation of the parasites at 37°C for 1 h caused a 70% increase in HSP70 mRNA which was prevented by ZC3H11 RNAi ( Figure 5F , lanes 3 & 4 ) . We transfected procyclic forms with the CAT reporters containing the full HSP70 3′-UTR , the HSP70 3′-UTR fragments or the actin 3′-UTR ( Supplementary Figure S4A ) and subjected the parasites to heat shock . This revealed that the AU-rich segment from the distal portion of the HSP70 3′-UTR was sufficient to confer persistence of the reporter mRNA in heat shock conditions ( Figure 5G , lanes 7–9 ) whereas the mRNA with the 5′ portion ( Figure 5G , lanes 4–6 ) behaved similarly to the actin control ( Figure 5G , lanes 10–12 ) . We concluded that ZC3H11 is required for the heat-shock response of procyclic trypanosomes , and that the heat-shock response is required for recovery of the parasites from incubation at 41°C . A previous microarray analysis had identified mRNAs that escape degradation after heat shock of procyclic forms [17] . We repeated this analysis by RNASeq , comparing the transcriptomes of procyclic trypanosomes after one hour at 41°C with those of parasites that remained at 27°C . A large number of mRNAs was affected ( Supplementary Table S1 , sheet 4 and Supplementary Figure S3D ) . In theory , the 41°C RNASeq data should be normalised to allow for the fact that the total amount of mRNA is four-fold diminished by heat shock ( Figure 5E ) , which means that the read count ratio ( heat shock/no heat shock ) should be divided by four . In practice , however , the un-normalised RNASeq results agreed better with those from Northern blots ( Table 1 ) . We do not understand why this is the case . Of the 178 loci that showed at least 2-fold more expression in the published heat shock microarray , 88 were confirmed as at least 2-fold increased in the RNASeq analysis; examples are listed in Table 2 . Intriguingly , the increased transcripts were significantly enriched for the class encoding RNA-binding proteins ( P = 0 . 007 ) . Several chaperones were actually decreased after heat shock , but these were preferentially those involved in vesicular transport ( Supplementary Table S1 , sheet 4 ) . Some of the mRNAs that increase after heat shock are also normally preferentially expressed in bloodstream forms ( Table 2 ) . Overall , there was no correlation between mRNA changes after heat shock and binding to ZC3H11 ( Supplementary Table S1 , sheet 4 ) , indicating that for most mRNAs , other regulatory mechanisms are involved in stabilisation after heat shock . To find mRNAs that were dependent on ZC3H11 after heat shock , we compared the transcriptomes of wild-type heat-shocked parasites with those of heat-shocked parasites with ZC3H11 RNAi . 27% of mRNAs were at least 2-fold less abundant in the RNAi cells; less than 1% were increased . Although there was no transcriptome-wide correlation between ZC3H11 binding and the RNAi effect ( Supplementary Table S1 , sheet 5 ) , every single one of the ZC3H11-bound heat-shock chaperone mRNAs was decreased in the RNAi cells ( Table 1 ) : the enrichment of chaperones in the subset that was both ZC3H11-bound and reduced in heat-shock was highly significant ( P = 1 . 6×10−13 ) ( Supplementary Table S1 , sheet 7 ) . The RNASeq results therefore showed that ZC3H11 is required for the retention of mRNAs encoding refolding chaperones after heat shock . The decreases in other mRNAs in the RNAi cells may be secondary to the loss of chaperones or other proteins encoded by ZC3H11-bound mRNAs . There are two basic ways in which an RNA-binding protein can stabilise an mRNA . One possibility is that it has a direct stabilising function , for example by binding to other proteins such as translation factors or poly ( A ) -binding protein . The other option is that it prevents binding of other , degradation-promoting , proteins to the recognition sequence . To distinguish between these possibilities , we artificially forced ZC3H11 to bind to a CAT reporter RNA that would not normally bind ZC3H11 . To do this we took advantage of the very strong and specific binding of the phage lambdaN protein to an RNA sequence called boxB . Briefly , we expressed a ZC3H11 protein fused to the peptide with binding activity ( λN-ZC3H11 ) ( Figure 6A ) together with a reporter RNA . The CAT reporter RNA was followed by five copies of the boxB sequence , then the actin 3′-UTR ( Figure 6A ) , and the same reporter mRNA without boxB served as a control . The effect of induction of the λN-ZC3H11 fusion protein ( Figure 6A , B ) on the CAT reporter mRNA and the protein level were examined . λN-ZC3H11 expression resulted in a 4-fold increase in reporter RNA , and a 2-fold increase in CAT protein ( Figure 6C ) . This effect required ZC3H11 binding , since the reporter lacking the boxB element did not react to ZC3H11 expression ( Figure 6C ) . We do not understand the difference between the protein and RNA effects; one possibility is that the tethered ZC3H11 has a negative influence on translation . We concluded that ZC3H11 is able to stabilise transcripts actively , rather than by inhibiting the action of a destabilising protein . Tethering the C-terminal part of ZC3H11 alone , without the zinc finger domain , also increased the reporter RNA and protein , while the N-terminal zinc-finger fragment did not ( Figure 6C ) . Together with our previous results , this suggests that the N-terminal zinc finger is required for RNA binding , while the effector domain for stabilisation lies towards the C-terminus .
Experiments with recombinant ZC3H11 fragments showed that , as expected , the zinc finger was able to interact with RNA , with little discrimination between AUU and AUUU repeats . With an RNA:protein ratio of 1∶100 ( Figure 2D ) , interaction with poly ( U ) was observed , but no binding was seen with ten-fold less RNA probe ( not shown ) and poly ( U ) competed extremely poorly with AUU repeats ( Figure 2E ) . This suggested that binding to poly ( U ) was weak . As for MEX5 [46] , no interaction was observed with poly ( C ) or poly ( A ) . The ability of ZC3H11 to discriminate against poly ( U ) was unexpected . In BRF2 , the glutamate of the ( R/K ) YKTEL motif is important for the specific interaction with the adenine base [31] . In ZC3H11 , lysine is found at this position - as in Caenorhabditis elegans MEX5 . MEX5 shows little discrimination between a classical 34mer ARE ( a mixture of ( AUU ) and ( AUUU ) repeats ) and U30; mutating the lysine to glutamate in both of the MEX5 zinc fingers to generate a more TTP/BRF2-like 6mer , ( N/K ) YKTEL , eliminated poly ( U ) binding [46] . The ability of Tis11 proteins to distinguish between different 3′-UTRs is dependent on the presence of two zinc finger motifs , resulting in a minimum 8-nt binding site of two UAUU motifs . The spacing between the A residues is defined by the distance between the two zinc fingers [31] . Monomeric ZC3H11 , in contrast , could recognise only UAUU so would have almost no ability to discriminate between mRNAs . The fact that the RNA pull-down yielded a highly specific motif of AUU repeats indicates that ZC3H11 must , in fact , bind at least as a dimer . The rather closer spacing of the A residues would , in that case , be determined by the geometry of the dimerization . Immunoprecipitations using V5-ZC3H11 co-expressed with ZC3H11-myc revealed no evidence for dimerization , although interference by the tags cannot be ruled out . It is possible that in vivo , interactions with other proteins serve to create ZC3H11 multimers . ZC3H11 is conserved throughout the Kinetoplastidae , as judged by reciprocal BLASTp analysis and alignments ( Supplementary Figure S1 ) . Nothing is known about the heat shock response in other salivarian trypanosomes , but some information is available for the more evolutionarily distant Trypanosoma cruzi . A proteome analysis found only minor changes after heat shock [47] , and the level of TcSTI1 was also unchanged , both at the steady state level and for polysomal RNA [48] . HSP70 mRNA is increased about two-fold after a 37°C heat shock of epimastigotes , and reporter experiments assigned responsibility to both 5′ and 3′-UTRs [49] . To see if the regulation by ZC3H11 might be conserved , we looked at the 3′-UTRs of the T . cruzi mRNAs encoding homologues of HSP70 ( Tb11 . 01 . 3110 ) , STI1 ( Tb927 . 5 . 2940 ) , DNAJ2 ( Tb927 . 2 . 5160 ) , HSP100 ( Tb927 . 2 . 5980 ) , FKBP ( Tb927 . 10 . 16100 ) , HSP110 ( Tb927 . 10 . 12710 ) and DNAJ1 ( Tb11 . 01 . 8750 ) ; all but the last one had extended ( AUU ) tracts . Thus at least as far as T . cruzi , the function of ZC3H11 might be conserved . The Leishmania ZC3H11 genes are not syntenic with those of trypanosomes , but nevertheless retain conserved sequence features ( Supplementary Figure S1 ) . Data for L . infantum indicate a 50% increase of ZC3H11 mRNA during differentiation to axenic amastigotes , but no change was seen in intracellular amastigotes of L . infantum or L . major ( http://tritrypdb . org/ ) . It has long been known that L . major HSP70 and HSP83 - like the trypanosome counterparts - are preferentially synthesised after heat shock [50] . 3′-UTR analysis , however , suggests that the sequences required are not the same . Studies of HSP70 in L . infantum showed that the 3′-UTR was responsible for mRNA stability after heat shock and that a region towards the 3′-end of it was required [24] , [25] . There are however no ( AUU ) repeats in the entire 3′-UTR . L . mexicana HSP70 ( LmxM . 28 . 2770 ) has UAUUUAUAUUAUAUU , but the function of this sequence is unknown . Similarly , the 3′-UTRs of the Leishmania HSP83 mRNAs are well conserved between species , but bear no resemblance to those of trypanosome HSP83s . The L . mexicana HSP83 mRNA lacks ( AUU ) repeats; a C/U-rich 150-nt region appears to be mainly responsible for regulation , perhaps via thermal melting [27] . The 3′-UTR of L . braziliensis HSP100 ( LbrM . 29 . 1350 ) , is devoid of ( AUU ) repeats , but L . mexicana HSP100 ( LmxM . 08_29 . 1270 ) has the classical ARE , UAUUUAUUUAUU . The lack of conservation of the L . mexicana AU-rich sequences suggests that any function would be species-specific . From this brief survey we suggest that ZC3H11 could be involved in regulating heat-shock genes in all trypanosomes . If it is involved in Leishmania , a different recognition sequence must be involved . After heat shock in procyclic forms , several mRNAs that bind to ZC3H11 are specifically stabilised . After an hour of heat shock , HSP70 and HSP83 mRNAs are also still actively translated ( Figure 5D and [17] , [18] ) . At the same time , heat-shock stress granules containing PABP and various translation initiation factors accumulate [17] . After 1 h heat shock , the amounts of phosphorylated ZC3H11 increased; only later , as cell viability decreased , did additional species appear which could either be dephosphorylated , or proteolytic fragments . We suggest that phosphorylation of ZC3H11 is increased after heat shock , resulting in increased ZC3H11 stability . The increase in ZC3H11 would stabilise existing HSP RNAs; since we saw no evidence for accumulation of ZC3H11 in granules , perhaps it prevents sequestration of target mRNAs . The mammalian zinc-finger proteins TTP and BRF1 both induce degradation of bound mRNA containing AU-rich elements , and both are recruited to stress granules and P-bodies [51] , [52] . Phosphorylation of TTP and BRF1 results in interaction with 14-3-3 isoforms , and consequent protein stabilisation , but at the same time , the functions of both factors in recruiting the mRNA degradation machinery , and their migration to stress granules , is impaired [51] , [53] , [54] . It is possible that phosphorylation of ZC3H11 acts in an analogous fashion . ZC3H11 RNAi killed bloodstream forms , but not procyclics . It is quite possible that in procyclics , ZC3H11 is indeed essential , but the residual ZC3H11 after RNAi was sufficient for survival . Notably , we were unable to delete both ZC3H11 genes although knockout constructs were able to integrate correctly into the genome ( not shown ) . The most likely explanation for the difference in RNAi behaviour is that procyclics are routinely grown at a lower temperature ( 27°C ) , with a lower chaperone requirement: ZC3H11 RNAi made procyclics hypersensitive to a temperature of 37°C . The top 120 different mRNAs enriched in the ZC3H11 bound fraction represent , in total , only 300 RNA molecules per bloodstream-form trypanosome [14] so a role for the rather low amounts of ZC3H11 in regulating target RNA abundance or translation is quite possible . However , the overall role of ZC3H11 in bloodstream forms is unclear , especially as results from RNASeq and Northern blots were not always concordant . Many bound mRNAs were not affected by the RNAi , but this is not really surprising . All trypanosome 3′-UTRs are long enough to interact with multiple RNA-binding proteins , which are expected to influence the behaviour of the bound mRNAs in a combinatorial fashion . Many RNAs that either increased or decreased after ZC3H11 RNAi were also not found bound to ZC3H11; perhaps the transcriptome after RNAi , at least in part , simply reflected the onset of growth inhibition . Further investigation - perhaps with transcriptomes taken at different times after ZC3H11 RNAi induction , and also comparing various other growth-inhibitory conditions - will be required to ascertain which effects of ZC3H11 depletion are direct . Chaperones assist in the folding of the nascent polypeptide chains , refold denatured proteins , and regulate the function of bound proteins . The T . brucei genome encodes large numbers of predicted or known chaperone-pathway proteins [55]: our domain searches revealed 5 Hsp70s , 11 Hsp20s , 35 petidyl-prolyl cis-trans isomerases , and 73 DnaJ-domain proteins . Among these , ZC3H11 mRNA binding was highly biased towards gene products that are required , not just for constitutive protein folding , but for recovery from stress . Notably , the complete refolding cycle [56] , [57] was represented: the major cytosolic HSP70; three DnaJ-domain proteins ( HSP40s ) and three Hsp20s , two TPR-repeat-containing peptidyl-prolyl cis-trans isomerases , the major cytosolic Hsp90 homologue HSP83 , HSP100 , HSP110 and the regulator STI1 . ZC3H11 bound mRNA encoding mitochondrial HSP60 , but not to any of those encoding the cytosolic TriC complex . ZC3H11 also bound to several mRNAs encoding proteins of unknown function , and without recognisable domains , some of which were elevated after heat shock , but are not required for normal growth [58] . We speculate that some of these could be involved in the heat-shock response . Since many of the chaperones that are involved in the trypanosome heat-shock response are essential at normal growth temperatures , it has hitherto not been possible to tell whether the heat-shock response itself ( as opposed to the constituent proteins ) has any role in Kinetoplastid survival . Indeed , in Leishmania , the steady-state levels of two of the major proteins , HSP83 and HSP70 , are so high that they are unaffected by the transient increases in their synthesis that occurs after heat shock [59] . We have now found that in procyclic forms , depletion of ZC3H11 not only prevented the continued synthesis of HSP70 and HSP83 , and the stabilisation of several other chaperone mRNAs , after a 41°C heat shock , but also severely impaired the ability of the trypanosomes to recover when returned to 27°C . This shows , for the first time , that a Kinetoplastid heat-shock response per se protects the parasite from short periods at an elevated temperature . At first sight , the temperatures that are required for the trypanosome heat shock response seem un-physiologically high . For bloodstream forms , an increase HSP70 mRNA was seen only at 41 . 5°C and above [60] . Such temperatures are rare in humans and many of the wild ungulates that are natural hosts for trypanosomes . Running African gazelles are , however , able to survive body temperatures of 43°C without ill-effects [6] , and trypanosomiasis in gazelles can cause fever temperatures of 43°C [7] . Procyclic trypanosomes show a heat-shock response at and above 37°C . Their host Tsetse flies are not found in environments in which the maximum air temperature exceeds 41°C [8]; Tsetse rest during the hottest part of the day [8] and feed preferentially on the lower ( more shady ) parts of animals [61] . Nevertheless , 10% of a population of Glossina pallipides were able to survive an hour at 41°C , and during gradual heating , some can survive up to 44°C; [62] . Overall , therefore , it seems that trypanosome heat-shock responses have evolved such that they are induced at 1–2°C below the maximum temperatures that are likely to be experienced in the relevant host . The heat-shock response is therefore tuned to enhance parasite survival under field conditions .
Details of all plasmids and oligonucleotides are provided in Supplementary Table S2 . All experiments were done with Lister 427 monomorphic procyclic or bloodstream form parasites expressing the Tet-repressor [63] . Procyclic forms were grown in MEM-Pros medium at 27°C ( unless stated otherwise ) at densities lower than 8×106 cells/ml . The bloodstream stage parasites were cultivated in HMI-9 medium in an incubator at 37°C with 5%CO2 at densities lower than 1 . 5×106 cells/ml . Additionally , various stable cell lines were created with constitutive ( CAT reporter/V5 ) or tetracycline-inducible expression ( RNAi , ectopic expression ) . For the tethering assays , cell lines constitutively expressing CAT reporter with actin 3′-UTR or boxB actin 3′-UTR were co-transfected with an inducible lambdaN-ZC3H11-myc fusion protein . Fragments of the ZC3H11 open reading frame ( first 104a . a . , 119a . a . , 136a . a . , 199a . a . , with or without a C→S mutation in the zinc finger ) were cloned into pQEA38 and expressed as His10-fusions in E . coli ( strain Rosetta , D3 pLysS , Novagen ) . The full-length ZC3H11 open reading frame was cloned in pET-trx1b ( pHD2222 ) and co-expressed with the groES-groEL-tig chaperone system ( Takara's Chaperone Plasmid , pG-Tf2 ) in E . coli ( strain Tuner , Novagen ) . Buffer in protein samples was exchanged to binding buffer ( 20 mM Tris-HCl , pH 8 . 0 , 50 mM NaCl , 100 µM ZnCl2 ) using Amicon Ultracentrifugal filter 10 kDA NMWCO columns . Protein aliquots were supplemented with 10% glycerol ( final concentration ) and stored at −80°C . Dephosphorylation assays [64] and co-immunoprecipitation assays [65] were done as previously described . Phosphatase inhibitors used were Sodium Orthovanadate ( 2 mM ) and Sodium Fluoride ( 8 mM ) . For co-immunoprecipitation with V5-ZC3H11 , 4×107 procyclic trypanosomes were pre-treated for 1 h with the proteasome inhibitor MG-132 ( Calbiochem ) at a concentration of 10 µg/ml . Cells were lysed in hypotonic buffer ( 10 mM NaCl , 10 mMTris-Cl pH 7 . 5 , 0 . 1% NP40 with complete protease inhibitor ( Roche ) , and precipitation was with anti-myc ( Biomol ) -coupled beads after the salt was adjusted to 150 mM NaCl . Proteins were detected by Western blotting . Antibodies used were to the V5 tag ( AbD seroTec , 1∶1000 ) , the Myc tag ( Santa Cruz Laboratories , 1∶1000 ) , aldolase ( rabbit , 1∶50000 [66] ) Detection was done using ECL solutions ( GE Healthcare ) . Chloramphenicol acetyltransferase was measured in a kinetic assay involving partition of 14C-buturyl chloramphenicol from the aqueous to the organic phase of scintillation fluid [67] . Trypanosomes were subjected to heat-shock at 41°C ( for 1 hour ) in a water bath and harvested immediately for RNA and Western blot . To measure protein synthesis , 2×106 cells were pelleted , resuspended in 500 µl of MEM lacking methionine . After 15 mins , [35S] methionine ( Amersham , 20 µCi ) was added and the cells were incubated at 27°C for 20 min . Pelleted cells were washed once ( 1× PBS+0 . 5%Glucose ) then resuspended in Laemmli sample buffer and subjected to SDS-PAGE . The gel was fixed in 10% acetic acid , 30% methanol solution in water for 45 min , stained with Coomassie followed by de-staining in water . The gel was then incubated for 45 min in En3Hance ( Amersham ) , washed in water for another 45 mins , dried and exposed for autoradiography . 2*108 procyclic cells were UV-cross-linked ( 400 mJ/cm2 ) prior to freezing in liquid nitrogen . Frozen pellets were resuspended in 500 µl hypotonic buffer ( 10 mM Tris pH 7 . 5; 10 mM NaCl; 0 . 1% IGEPAL ) containing protease inhibitor ( complete mini EDTA free; Roche ) and 8 mM VRCs ( Sigma ) and 800 u RNasin ( Promega ) . Lysis was completed by passing 15 times through a 27G needle . After pelleting insoluble debris ( at 6000 g ) and adjusting to 150 mM NaCl , the protein was allowed to bind for one hour at 4°C to anti-myc coupled agarose ( Biomol ) . The agarose was then washed 4 times at 4°C with IPP150 ( 10 mM Tris pH 7 . 5; 150 mM NaCl; 0 . 1% IGEPAL ) . Cross-linked protein was digested with 20 µg proteinase K at 42°C for 15 min , and RNA was isolated from both pellet and unbound fractions using Trifast ( Peqlab ) . Total RNA was extracted using Trifast ( Peqlab ) . Blotted RNA was detected by hybridization with radioactive probes ( see Supplemental Table S2 ) . For microarray analysis of cells with ZC3H11 RNAi , we used oligonucleotide arrays ( mostly one oligonucleotide per open reading frame ) from NIAID and TIGR . They were hybridised with cDNA made from total RNA from either bloodstream forms expressing the repressor , or cells with RNAi after one day of tetracycline induction , as previously described [68] . For analysis of RNA bound to ZC3H11 , the unbound RNA fraction was treated with the Ribominus kit ( Invitrogen ) to remove ribosomal RNAs . Both eluate and unbound RNAs were then subjected to deep sequencing , and reads aligned to the genome as previously described [14] . To normalise the data we calculated the total number of reads in a set of unique genes [69] , with the major HSP70 , GPEET procyclin and EP procyclin added . This number was used in order to calculate the reads per million total reads . To extract all known 3′-UTRs , we used all of our available sequence sets and extracted the reads that failed to align to the genome . From these , we selected those that contained at least five contiguous T residues at the start of the read . After removal of the T's , we re- aligned the reads to the genome . For each ORF , the most abundantly aligned sequence was assigned as the major polyadenylation site; if two gave equal read densities , the proximal one was used . The output is available on request . For mRNAs that were classified as strongly ( 3× ) enriched in the ZC3H11-bound sample , whenever 3′-UTRs were missing or too short we extracted them manually using published information and processing sites annotated in the TritrypDB database , or simply used 200 bp downstream of an ORF . MDscan [70] was used to search for motifs in the 3′-UTRs of mRNAs that were at least 3-fold enriched in the eluate , using the top 10 bound mRNA's as seeds . A similar motif search was also done using MotifSampler [71] with a 3rd order background model constructed from all known 3′-UTRs . The motif with the highest representation in the MotifSampler output , and highest log likelihood score , was identified as a potential binding motif of ZC3H11 , and it was in agreement with the MDscan motif . Next , FIMO [72] was used to scan all known 3′-UTRs using the model of the motif we found , and only hits with a q-value<0 . 01 were retained . RNASeq analysis was performed on poly ( A ) + mRNA , except for the procyclic wild-type sample , which was part of another study and was rRNA depleted , not poly ( A ) selected . Previous results with bloodstream forms indicated that at steady state , these populations are similar ( R = 0 . 96 ) [14] . The RNA was fragmented before cDNA synthesis , using the standard Illumina protocol . Alignment and normalisation were as described above . The relevant Supplementary Table S1 sheets ( 3–5 ) include all unique genes with rpm of at least 10 in each of the relevant datasets . Since these experiments were performed only once , we focus in our discussions on results that were also verified by other means . Oligoribonucleotides ( 50 pmol , Biomers , 50 µl reaction ) were 5′-end labelled with 50 µCi of gamma-[32P] ATP and 20 units of T4 polynucleotide kinase ( New England Biolabs ) for 37°C for 30 min , then purified using the QIAquick Nucleotide Removal Kit ( QIAGEN ) . Recombinant proteins ( 100 pmol ) were incubated for 30 min at room temperature in binding buffer ( 20 mM Tris-HCl , pH 8 . 0 , 50 mM NaCl , 10 µM ZnCl2 , 0 , 01% IGEPAL CA-630 , 0 . 1 mg/ml of tRNA , 10 µg/ml of heparin ) with 1 pmol radiolabeled RNA , with or without 5- and 20-fold excess competitor RNA [73] . Loading dye was added and the samples were run on 5% non-denaturing polyacrylamide gels , ( 0 . 3× tris-borate buffer +100 µM ZnCl2; 30 min at 300 V ) which were then dried and analysed using a phosphorimager . For quantitative analyses , varying concentrations of protein were incubated with 10 pmol labelled RNA [73] .
|
When organisms are placed at a temperature that is higher than normal , their proteins start to unfold . The organisms protect themselves by increasing the synthesis of “heat-shock” proteins which can re-fold other proteins when the temperature returns to normal . In trypanosomes , the degradation of mRNAs that encode heat-shock proteins is slowed down at elevated temperatures . Trypanosoma brucei multiplies as “bloodstream forms” in the blood of mammals , at temperatures between 37–39°C; and as “procyclic forms” in Tsetse flies , which are usually at 20–37°C but can survive at 41°C . In this paper we show that in Trypanosoma brucei , a protein called ZC3H11 can bind to many heat-shock-protein mRNAs . ZC3H11 is essential in bloodstream-form trypanosomes and for recovery of procyclic-form trypanosomes after heat shock . ZC3H11 binds to an AUU repeat motif which is found in parts of the target mRNAs that do not encode protein . Several heat-shock-protein RNAs were decreased when we decreased the amount of ZC3H11 in bloodstream-form trypanosomes . These and other results show that expression of the specific subset of trypanosome heat-shock proteins is controlled by the interaction of ZC3H11 with the relevant mRNAs . They also show that the heat-shock response could enhance survival of trypanosomes in over-heated Tsetse flies .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"rna",
"cellular",
"stress",
"responses",
"molecular",
"cell",
"biology",
"cell",
"biology",
"nucleic",
"acids",
"biology",
"microbiology",
"rna",
"stability",
"molecular",
"biology",
"parasitology",
"parasite",
"physiology"
] |
2013
|
Post-Transcriptional Regulation of the Trypanosome Heat Shock Response by a Zinc Finger Protein
|
The genealogical relationship of human , chimpanzee , and gorilla varies along the genome . We develop a hidden Markov model ( HMM ) that incorporates this variation and relate the model parameters to population genetics quantities such as speciation times and ancestral population sizes . Our HMM is an analytically tractable approximation to the coalescent process with recombination , and in simulations we see no apparent bias in the HMM estimates . We apply the HMM to four autosomal contiguous human–chimp–gorilla–orangutan alignments comprising a total of 1 . 9 million base pairs . We find a very recent speciation time of human–chimp ( 4 . 1 ± 0 . 4 million years ) , and fairly large ancestral effective population sizes ( 65 , 000 ± 30 , 000 for the human–chimp ancestor and 45 , 000 ± 10 , 000 for the human–chimp–gorilla ancestor ) . Furthermore , around 50% of the human genome coalesces with chimpanzee after speciation with gorilla . We also consider 250 , 000 base pairs of X-chromosome alignments and find an effective population size much smaller than 75% of the autosomal effective population sizes . Finally , we find that the rate of transitions between different genealogies correlates well with the region-wide present-day human recombination rate , but does not correlate with the fine-scale recombination rates and recombination hot spots , suggesting that the latter are evolutionarily transient .
The recent evolutionary history of the human species can be investigated by comparative approaches using the genomes of the great apes: chimpanzee , gorilla , and orangutan [1] . Nucleotide differences , accumulated by fixation of mutations , carry a wealth of information on important issues such as speciation times , properties of ancestral species ( e . g . , population sizes ) , and how speciation occurred [2–5] . Genes or genomic fragments with unusual patterns of nucleotide differences and divergence may have been under strong natural selection during recent evolution of the human species [6] . Sequence analyses can also aid interpretations of the incomplete primate fossil records [7] and aid assignment of dated fossils to evolutionary lineages . For instance , it is still debated whether the Millennium man , Orrorin tugenensis [8 , 9] , which has been dated to 6 million years ( Myr ) ago , and Sahelanthropus tchadensis [10–12] , which has been dated to 6–7 Myr ago , belong to the human lineage or the human–chimp ( HC ) lineage . Comparative analyses of multiple alignments of small fragments of human , chimpanzee , gorilla , and orangutan sequence have revealed that the human genome is more similar to the gorilla genome than to the chimpanzee genome for a considerable fraction of single genes [2 , 13–15] . Such a conflict between species and gene genealogy is expected if the time span between speciation events is small measured in the number of 2N generations , where N is the effective population of the ancestral species ( see Figure 1 ) . In that case , N can be estimated from the proportion of divergent genealogies if one assumes that speciation is an instantaneous event . Indeed , this has been done in several studies that find a HC ancestral effective population size NHC of 2–10 times the human present-day effective population size NH = 10 , 000 [13 , 14 , 16–18] . Recently , Patterson et al . [2] studied a very large number of small human–chimp–gorilla–orangutan–macaque alignments . They found , in agreement with O'hUigin et al . [15] , that a large proportion of sites supporting alternative genealogies are caused by hypermutability and that the fraction of the genome with alternative genealogies therefore has been overestimated in previous studies . After using a statistical correction for substitution rate heterogeneity , Patterson et al . [2] found that the variance in coalescence times is too large to be accounted for by instant speciation and a large ancestral effective population size , and that the speciation process therefore must have been complex . Particularly , the X chromosome shows a deviant pattern , which also led them to conclude that HC gene flow ceased and final speciation occurred as recently as 4 Myr ago . This date is generally believed to be the most recent time compatible with the fossil record , if the Millennium man and Sahelanthropus are not on the human lineage . Whole genome sequences of gorilla and orangutan will soon supplement the already available whole genome sequences of human and chimpanzee [19] . These four genomes are so closely related that alignments of large contiguous parts of the genomes can be constructed . Analysis of such large fragments is challenging because different parts of the alignment will have different evolutionary histories ( and thus different genealogies , see Figure 1 ) because of recombination [14 , 20] . Ideally , one would like to infer the genealogical changes directly from the data and then analyze each type of genealogy separately . A natural approach to this challenge is to move along the alignment , and simultaneously compute the probabilities of different relationships and speciation times . While recombination has been considered in previous likelihood models [14] , the spatial information along the alignment has largely been ignored . In this paper we describe a hidden Markov model ( HMM ) that allows the presence of different genealogies along large multiple alignments . The hidden states are different possible genealogies ( labeled HC1 , HC2 , HG , and CG in Figures 1 and 2 ) . Parameters of the HMM include population genetics parameters such as the HC and human–chimp–gorilla ( HCG ) ancestral effective population sizes , NHC and NHCG , and speciation times τ1 and τ2 ( see Figure 1 ) . We therefore name our approach a coalescent HMM ( coal-HMM ) . The statistical framework of HMMs yields parameter estimates with associated standard errors , and posterior probabilities of hidden states [21–23] . We show by simulation studies that the coal-HMM recovers parameters from the coalescence with recombination process , and we apply the coal-HMM to five long contiguous human–chimp–gorilla–orangutan ( HCGO ) alignments obtained from the NIH Intramural Sequencing Center comparative sequencing program ( Targets 1 , 106 , 121 , and 122 on four different autosomes and Target 46 on the X chromosome ) . We consistently find very recent estimates of HC speciation times and a large variance in the time to common ancestry along the genome . Similar to Patterson et al . [2] , we find that the X chromosome has a smaller effective population size than expected . The mapping of genealogical states further allows us to correlate transitions in genealogies with properties of the genome , and here we focus on fine-scale [24] and region-wide [25] recombination rate estimates .
A visual impression of the preferred phylogenetic state along the alignment is obtained by dividing the alignment sites according to how they partition the species ( Table 1 ) . For example , sites where human and gorilla have the same base pair , which is different from chimpanzee , suggest a human–gorilla ( HG ) grouping . The HG grouping is further supported if the outgroup ( orangutan ) has the same base pair as the chimpanzee . In Figure 3 , we show the first 100 kb from Target 1 . The preferred topologies along the alignment are shown in the third ( support with outgroup ) and fourth ( support without outgroup ) panels of the figure , from top to bottom . The posterior probabilities of the phylogenetic states are shown in the panel second from the top in Figure 3 . Similar figures for all targets in 100-kb blocks can be found in Figures S1–S5 . The density and character of the strongly informative sites in the third panel and singletons in the fourth panel largely determine the inferred states along the alignment . State HC1 is generally the preferred state; this state is often strongly supported over contiguous portions of the alignment and typically spans several kilobases . The alternative states HC2 , HG , and CG ( chimp–gorilla ) are less strongly supported and typically cover very short segments . The upper panel shows the fine-scale recombination rates determined from human polymorphism data [24] . No clear association is observed between transitions in the HMM and these rates , e . g . , recombination hotspots are not concentrated in regions of rapidly changing genealogies ( for a formal statistical test , see Text S1 ) . Parameter estimates with standard errors for each target are shown in Figure 4 . Assuming orangutan divergence 18 Myr ago [18] , speciation time of human and chimpanzee is consistently around 4 Myr with small standard errors except for the very short Target 122 . The speciation times for the X chromosome are also in close agreement with the speciation times of the autosomes . The orangutan divergence assumed is the molecular divergence time , which may be much larger than the orangutan speciation time if the NHCGO effective population size was large . If another orangutan divergence time Z is preferred , then all our time estimates should be multiplied by Z/18 . The divergence times in Figure 4 are much higher than speciation times because of large effective population sizes in the ancestral species . We note that the effective population size of the HCG ancestor is more accurately determined than that of the HC ancestor , which may seem counterintuitive , suggesting that ancestral inference of certain quantities does not necessarily become increasingly difficult further back in time . Exactly the same pattern was found in simulations , suggesting that this is a true property of the process . Target 1 was also analyzed after filtering out all putative CpG mutations . This reduces the number of polymorphic sites by 17% and removes relatively more of the sites supporting HG and CG groupings than sites supporting HC grouping ( states HC1 and HC2 ) , as expected if some of these sites are hypermutable . However , the removal of putative CpG sites does not change the estimated time in alternative states or effective population sizes and only slightly decreases the estimated HC speciation time . The time spent in the alternative states HC2 , HG , and CG is also only slightly affected ( Text S2 ) . The visual impression of figures such as Figure 3 also remains the same after filtering out the hypermutable CpG sites . We also aligned Target 1 to a further outgroup , gibbon , in order to identify sites showing evidence for recurrent mutations resulting in implausible five-species site patterns ( see also [2] ) . Removing these sites further reduces the proportion of strongly informative sites supporting the HG and CG states , but the proportion of time spent in each state , speciation times , and divergence time estimates are again only slightly affected after removal of these sites ( see Text S2 ) . While estimates of the effective population sizes and speciation times do not differ significantly between the four targets , there are large differences in the average number of base pairs and the percentage of the alignment in state HC1 ( Table 2 ) . The average number of base pairs in state HC1 correlates well with the average recombination rate of the target estimated from pedigree data [25] . Thus , this broader scale recombination rate appears to be conserved over a longer time scale than the fine-scale recombination rate estimated from human diversity data , and it appears to be shared with chimpanzee and gorilla . A coal-HMM analysis of more than 250 kb of X-chromosome sequence data used by [2] shows that 78% ± 5% of the alignment supports state HC1 ( Figure 2 ) . One of the deviant regions is shown in Figure 5 , where a cluster of sites supporting a HG relationship is observed ( figures for the whole target are available in Figure S5 ) . Thus we find that not all of the X-chromosome data are consistent with state HC1 as previously suggested [2] . However , only a small fraction of the X-chromosome data support alternative states , and this is consistent with an effective population size of the X chromosome in the HC ancestor of approximately 17 , 000 ( see Figure 4 ) , which is much lower than expected assuming an effective population size of 75% that of the autosomes . Curiously , the HCG ancestor NHCG for the X chromosome data is close to the expected ( see Figure 4 ) .
Studying the genealogical relationship of human , chimpanzee , and gorilla along their genomes makes it possible to assign genealogical relationships to segments of the genome with high posterior probabilities ( Figures 3 and S1–S5 ) . Long fragments of several kilobases supporting the basic state HC1 alternate with kilobase-long fragments that support the alternative states HC2 , HG , and CG . Since alternative states imply coalescence further back in time , the ancestral material is expected to be broken up more by recombination in regions supporting these states . Thus , frequent changes among alternative states are predicted by coalescent theory , but this has usually not been explicitly considered in previous analyses . The picture of alternating genealogies can subsequently be correlated with genomic features such as specific genes suspected to be important in human evolution , and can be used to survey whole genomes for extraordinarily long segments indicative of selection and/or recent introgression . The complex speciation model suggested in Patterson et al . [2] can also be more closely investigated using extensions of our coal-HMM framework when longer contiguous alignments become available . Important insights can already be gained from analyzing 1 . 9 million base pairs from four autosomal segments of the genome , and 0 . 25 million base pairs from the X chromosome . We consistently find that the speciation time of human and chimpanzee is close to the minimum of the range previously predicted ( 4 Myr ) if we assume a human–orangutan divergence of 18 Myr . If the effective population size NHCGO was large , then there is also a large variation in orangutan divergence . However , 18 Myr ago , the size of ancestral segments was very short ( a few base pairs ) , so the variation in divergence times over kilobase-long fragments as studied here is expected to be small . The HCG speciation time is estimated to have occurred approximately 2 Myr earlier than the HC speciation time ( i . e . , 6 Myr ago ) . However , the divergences along the genome of human and chimpanzees are generally much older than 4 Myr , varying between 4 and 9 Myr . This can be explained solely by ancestral population sizes on the order of 50 , 000 in the HC and HCG ancestors , and one does not need to invoke a gradual speciation process with continued gene flow ( introgression ) to explain the autosomal data , as also noted by Innan and Watanabe [4] and Barton [5] . Our molecular dating estimates are generally in agreement with a large number of studies using different calibration points; Kumar et al . [26] , Glazko and Nei [27] , and even the classical study of Sarich and Wilson [28] found a molecular divergence of HC at 5–7 Myr , 6 Myr , and 5 Myr , respectively . Speciation , defined as the total cessation of gene flow , is necessarily more recent than these molecular dates , and our value of approximately 4 Myr agrees very well with the time suggested by Patterson et al . [2] for complete cessation of gene flow . It is also in agreement with the oldest fossils generally accepted to belong to the human lineage after the HC split . The autosomal analysis alone cannot be used to determine if the large variance in coalescence times of human and chimp along the genome is due to a large ancestral effective population size or due to prolonged speciation [5] . The present implementation of the coal-HMM assumes that , conditional on the genealogy , sites are independent and mutations can be described by a continuous time Markov chain . This assumption is violated for CpG dinucleotides , which are more prone to mutation due to methylation . The assumption is also violated if the mutation rate varies along the alignment , resulting in regions with highly variable sites that are subject to recurrent mutations . Recurrent mutations can be detected by adding a further outgroup [2 , 15] . We used a similar strategy as Patterson et al . [2] to mask CpG-induced variable sites and sites with clear evidence of recurrent mutation on the human–chimp–gorilla–orangutan–gibbon phylogeny . Our analyses show that removal of these sites does not affect our estimates of genealogies along the alignment , of the divergence times , or of the speciation times very much . We ascribe this robustness of the coal-HMM to the fact that the model uses information from the singleton sites , which are much more common than the strongly informative sites ( recall Table 1 ) . Indeed , we have calculated that the evidence for distinguishing between genealogical states of a strongly informative site corresponds to 5–7 singletons supporting the same state . The 250 , 000 base pairs aligned on the X chromosome have a larger than expected fraction of base pairs in state HC1 ( around 80% ) if the long-term effective population size of the X chromosome is three-quarters that of the autosomes . There is strong evidence that a small segment on the X chromosome supports an alternative genealogy ( state HG ) . However , the effective population size is reduced by much more than the expected 25% assuming instant speciation , equal contribution of sexes , and no selection . Non-equal contribution of sexes can at most reduce the effective population size to 50% that of autosomes ( if females have much larger variance in reproductive success than males ) . Prolonged speciation and selection may both explain the discrepancy . The X chromosome , with its hemizygosity in males , is more exposed to selection , and this can be an argument for a different introgression history ( in a prolonged speciation process ) or for selection generally affecting the X chromosome more in the HC ancestor , thus reducing the effective population size more than on the autosomes . The observed fraction of 80% in state HC1 is expected when the effective population size of the X chromosome is approximately 35% that of the autosomes . Explicit modeling of introgression processes and the natural selection on the X chromosome , together with an extended coal-HMM ( that explicitly models the length distribution of segments supporting different genealogies ) , may provide the means to test among these alternative explanations when more data become available . Another feature of the human genome that can be explored using the coal-HMM is the evolutionary scale of variation in recombination rate . We observe a correlation between the region-wide recombination rate estimated from pedigrees and the average length of segments in state HC1 , consistent with evolutionary conservation of these region-wide recombination rates . However , we see no clear correlation between the fine-scale recombination rate estimated from human polymorphism data and transitions to and from state HC1 . We interpret this as additional evidence that recombination hot spots are very transient , since we are analyzing an even shorter time scale ( at least half ) than that used when comparing human and chimpanzee , where recombination hot spots do not appear to be shared [24] . When more data become available , the coal-HMM can be extended to investigate more specific hypotheses . Our simulations show that the coal-HMM provides a reasonable approximation to the much more complex coalescent-with-recombination process . Furthermore , we found that adding more parameters in the transition probability matrix did not improve the fit significantly with 1 . 9 million base pairs . With a 1 , 000-fold increase in data soon to arrive , a natural extension is to introduce more hidden states in the HMM to provide a more detailed approximation of the different coalescent times . When this is done we expect a better fit of segment sizes to the geometric distribution assumed by the coal-HMM . Importantly , having more data will make it possible to investigate changes in ancestral effective population sizes along the genome , thus making it possible to infer cases of ancestral selection in the HC ancestor and in the HCG ancestor , and opening up a promising field of ancestral species population genetics that can complement analyses based on dn/ds ratios . It will also be possible to explicitly test whether the effective population sizes of ancestral species NHC and NHCG have changed through time .
Coal-HMMs provide the framework for analyses of genome alignments of human , gorilla , chimpanzee , and orangutan sequences . Coal-HMMs are similar to phylogenetic HMMs [29] , but instead of partitioning the alignment into fragments undergoing different evolutionary processes because of functional properties ( e . g . , noncoding , exonic , and intronic regions ) , the alignment is partitioned into fragments of different evolutionary histories separated by recombination events . In our coal-HMM we consider recombination events that separate four different genealogies . The four genealogies are shown in Figure 1 and correspond to the hidden states of the model . The transitions between the hidden states are modeled using a Markov chain with transition probability matrix P ( · , · ) . We have primarily studied a transition matrix given by The stationary distribution of the Markov chain is where ψ = 1/ ( 1 + 3s/u ) . The initial state probability of the coal-HMM is given by Ψ . We also investigated more parameter-rich transition probability matrices . In particular , we considered a symmetric model for the transitions between the HC2 , HG , and CG states with three parameters ν1 , ν2 , and ν3 , where ν1 is the probability of making a transition between HC2 and HG , ν2 the transition probability between HC2 and CG , and ν3 the transition probability between HG and CG . However , such extended models did not improve the fit significantly with the present amount of available data . Let X = [X1 , … , XL] denote the alignment , consisting of L columns ( sites ) and four rows ( corresponding to the four species ) . The probability Pe ( Xi | φi ) that an alignment column Xi is emitted from the hidden state φi ∈ {HC1 , HC2 , HG , CG} is determined by the phylogenetic tree corresponding to the hidden state and a substitution rate matrix Q . We considered several rate matrices and found that the strand-symmetric rate matrix ( e . g . , [30] ) provided a good description of the data . This is perhaps not surprising because the data we analyzed primarily consist of noncoding sequences . The strand-symmetric substitution process has stationary frequencies π = ( πA , πG , πC , πT ) , where πA = πT and πG = πC . We calibrate the rate matrix such that branch length corresponds to expected substitutions per site . The branch lengths are a , b , and c in state HC1 and a~ , b~ , and c~ in states HC2 , HG , and CG ( see Figure 6 ) . We would like to emphasize that continuous time Markov chains take recurrent mutations into account . Furthermore , the so-called CpG effect ( higher mutation rates from CpG → TpG and CpG → CpA on the opposite strand ) is also taken partially into account because C → T and G → A have particularly high rates in our estimated strand-symmetric rate matrix . For more information on recurrent mutations and CpG hypermutability , refer to Text S2 . Let η denote the free parameters in the coal-HMM such that η determines the transition matrix P ( φi-1 , φi ) = Pη ( φi-1 , φi ) , initial state probability vector Ψ = Ψη , and emission probabilities . The joint probability of an alignment X and a segmentation φ = ( φ1 , … , φL ) of the genealogies is then given by The likelihood is the sum over all possible segmentations . In the next subsection , we derive the free parameters η in the coal-HMM from the coalescent process with recombination . The derivation of the relation between parameters in the coal-HMM and the coalescent process with recombination is carried out in two steps , corresponding to the left and right illustrations in Figure 6 . We refer the reader to Chapter 5 in [20] for a thorough description of the coalescence process with recombination and to Yang [16] for a very similar derivation . In the inference procedure we measure all times in expected number of substitutions per site , and then subsequently rescale using an 18-Myr human–orangutan divergence time . First , consider the situation in the left half of Figure 6 . The parameters in the coalescent process are the coalescence time THC of two given lineages in the HC ancestor , and the coalescence time THCG of two lineages in the HCG ancestor . The two coalescence times THC and THCG are independent and exponentially distributed with means θHC = 2NHCμ and θHCG = 2NHCGμ , where NHC and NHCG are the effective population sizes in the HC and HCG ancestral populations and μ is the mutation rate . The probability that a randomly chosen site belongs to state HC1 is given by providing a relation between ψ and τ2/θHC . From this observation we also obtain and therefore the average coalescence time for human and chimpanzee in state HC1 is given by In state HC1 , we therefore obtain the following relations between the coal-HMM parameters ( a , b ) and the coalescent-with-recombination parameters ( τ1 , τ2 , θHC , θHCG ) : Second , we consider states HC2 , HG , and CG . The situation is depicted in the right part of Figure 6 for state HC2 . Conditioning on THC >τ2 , the coalescence of any two of the human , chimpanzee , and gorilla sequences is equally likely , and therefore we obtain similar equations as below for states HG and CG . Let THCG , 3 be the time to coalescence of any two given lineages when three lineages are present in the HCG ancestor . Standard coalescent theory says that THCG , 3 is exponentially distributed with mean θHCG/3 . We now obtain the following two equations Note that there are five parameters in the coal-HMM , but only four coalescent parameters ( τ1 , τ2 , θHC , θHCG ) in equations 1–5 . Thus , there is a constraint on the parameters in the coal-HMM . Subtracting equations 4 and 5 we get and subtracting equations 3 and 4 we get We thus obtain the constraint To identify the parameters we solve the system of equations 1–3 and 5 . From equation 1 we obtain Subtracting equation 3 from 2 and substituting the above we see that and therefore the parameters in the coalescence process are given by When reporting the parameters , we scale ( a , b , c ) such that ( a + b + c ) sums to twice the divergence time between human and orangutan , which is set to 18 Myr . We find the ancestral population sizes by assuming a generation time of 25 y . We assume the branch lengths fulfil the molecular clock constraint and the relation in equation 6 above . Thus the seven free parameters in the coal-HMM are and the maximum likelihood estimates are found using the Baum-Welch method ( e . g . , [31] ) . Standard errors for the free parameters are determined from the numerically evaluated Fisher information matrix . Standard errors for functions of the parameters are found using the delta method [32] . In order to validate the coal-HMM approximation to the coalescent process with recombination , we conducted a simulation study . The parameters in the simulated coalescent process with recombination are NHC = NHCG = 40 , 000 , NH = NC = NG = 30 , 000 , speciation time τ1 of HC is 4 Myr , speciation time τ2 of HCG is 5 . 5 Myr , the time to the ancestor of all four sequences is 18 Myr , the generation time of individuals is 25 y , the length of the sequence is 500 , 000 bp , and the recombination rate is r = 0 . 0075 ( corresponding to a genetic recombination frequency of 1 . 5 cM per Mb ) . The simulation from the coalescent-with-recombination process results in a number of recombination events , some of which are visible as change-points along the sequence where the phylogenetic tree changes . We then obtained a sequence alignment where for each position we simulated the evolution of a nucleotide on the phylogenetic tree at that position , and according to the strand-symmetric substitution process . The substitution rate was chosen to match the typical branch lengths in the HCGO quartet , 0 . 1% change per million years . Table 3 shows means and standard deviations of the most important quantities for 20 independent simulations . All main quantities are estimated without strong bias . Curiously , NHC is estimated with much larger variance than NHCG , in agreement with analyses of the real data . One of the assumptions of the coal-HMM is that the distribution of fragment lengths for each hidden state is geometric . Considering simulations from a coalescence with recombination process , the fragments for the coal-HMM are aggregations of fragments with branch lengths corresponding to the same state of the coal-HMM . Figure 7 shows the distribution of aggregated fragment sizes from a coalescent-with-recombination simulation . Although the coalescent-with-recombination process is non-Markovian when viewed as a process along the sequence [33] , the fragment lengths are reasonably well-described by a geometric distribution . The slight deficiency of very short fragments in Figure 7 is due to the aggregation of fragments with correlated coalescence times that occurs for each state . Figure 8 compares simulated and inferred genealogies from a segment of 100 , 000 base pairs from one of the simulation runs . Coloring corresponds to the different states of the Markov chain ( see Table 1 ) , and the upper panel shows the posterior probability of each of the states . We see long fragments where state HC1 is the true state , and we see regions where the true state changes frequently . This is caused by the coalescence time for states HC2 , HG , and CG being further back in time than the coalescence time for state HC1 , leaving more time for recombination events to accumulate and cause more frequent changes of genealogy . This phenomenon is also reflected in the reconstruction , i . e . , in the posterior probabilities , where we see good agreement between true and inferred genealogical states of the long fragments where state HC1 is the true state , and we also see that areas where the true states change frequently are inferred as such . However , in addition , we see that within regions where the true states change frequently , the reconstruction in general is quite uncertain . Chimpanzee–gorilla–orangutan sequence data from Targets 1 ( Chromosome 7 ) , 106 ( Chromosome 20 ) , 121 ( Chromosome 2 ) , and 122 ( Chromosome 20 ) were obtained from the NIH Intramural Sequencing Center Web site ( http://www . nisc . nih . gov ) , and the corresponding human sequences were downloaded from GenBank ( http://www . ncbi . nlm . nih . gov/Genbank ) . For each target , sequences from the four species were aligned using the MAVID alignment software [34] . The resulting alignment was manually inspected , and columns corresponding to an insertion in orangutan were removed . Next , the alignment was subdivided in cases where more than 50 deletions were present in one of the species , and finally , alignment columns with gaps were removed . For the X-chromosome , the alignment used by [2] was downloaded and filtered for poorly aligned segments ( more than five segregating sites in 20 base pair windows ) . Fine-scale ( from HapMap data [24] ) and region-wide ( from pedigree data [25] ) estimates of the recombination rate were downloaded from the University of California Santa Cruz genome browser ( http://www . genome . ucsc . edu ) .
|
Primate evolution is a central topic in biology and much information can be obtained from DNA sequence data . A key parameter is the time “when we became human , ” i . e . , the time in the past when descendents of the human–chimp ancestor split into human and chimpanzee . Other important parameters are the time in the past when descendents of the human–chimp–gorilla ancestor split into descendents of the human–chimp ancestor and the gorilla ancestor , and population sizes of the human–chimp and human–chimp–gorilla ancestors . To estimate speciation times and ancestral population sizes we have developed a new methodology that explicitly utilizes the spatial information in contiguous genome alignments . Furthermore , we have applied this methodology to four long autosomal human–chimp–gorilla–orangutan alignments and estimated a very recent speciation time of human and chimp ( around 4 million years ) and ancestral population sizes much larger than the present-day human effective population size . We also analyzed X-chromosome sequence data and found that the X chromosome has experienced a different history from that of autosomes , possibly because of selection .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"primates",
"genetics",
"and",
"genomics"
] |
2007
|
Genomic Relationships and Speciation Times of Human, Chimpanzee, and Gorilla Inferred from a Coalescent Hidden Markov Model
|
Borrelia burgdorferi , the causative agent of Lyme disease in humans , is maintained in a complex biphasic life cycle , which alternates between tick and vertebrate hosts . To successfully survive and complete its enzootic cycle , B . burgdorferi adapts to diverse hosts by regulating genes required for survival in specific environments . Here we describe the first ever use of transposon insertion sequencing ( Tn-seq ) to identify genes required for B . burgdorferi survival in its tick host . We found that insertions into 46 genes resulted in a complete loss of recovery of mutants from larval Ixodes ticks . Insertions in an additional 56 genes resulted in a >90% decrease in fitness . The screen identified both previously known and new genes important for larval tick survival . Almost half of the genes required for survival in the tick encode proteins of unknown function , while a significant portion ( over 20% ) encode membrane-associated proteins or lipoproteins . We validated the results of the screen for five Tn mutants by performing individual competition assays using mutant and complemented strains . To better understand the role of one of these genes in tick survival , we conducted mechanistic studies of bb0017 , a gene previously shown to be required for resistance against oxidative stress . In this study we show that BB0017 affects the regulation of key borrelial virulence determinants . The application of Tn-seq to in vivo screening of B . burgdorferi in its natural vector is a powerful tool that can be used to address many different aspects of the host pathogen interaction .
Lyme disease is caused by the spirochete , Borrelia burgdorferi . In nature , B . burgdorferi is maintained in a cycle between mammalian or bird hosts and Ixodes ticks [1] Newly hatched ticks can acquire B . burgdorferi from infected animals during their larval feeding [1] . After molting to the nymphal stage , those infected ticks can transmit the pathogen to a new vertebrate host during their next blood meal [1] . The challenges posed by the vertebrate and tick environments are quite different . B . burgdorferi must adapt to changes in temperature , pH , nutrient availability and immune defense mechanisms between its vertebrate and arthropod hosts [2–6] . Previous studies have shown that B . burgdorferi adapts to its host environments through controlling the expression of proteins that aid in survival at specific points in its life cycle in its different hosts [7–9] . For example , proteins such as outer surface protein C ( OspC ) , variable-major- protein ( Vmp ) -like sequence E ( VlsE ) and decorin binding protein A ( DbpA ) are expressed to differing amounts during particular time points in the mammalian and tick phases of the B . burgdorferi life cycle [10–14] . The regulation of gene expression in B . burgdorferi is complex , often involving multiple layers of control [1 , 3 , 6] . Expression of proteins required during the mammalian phase involves two alternative sigma factors , RpoS and RpoN , the enhancer binding protein Rrp2 , as well as the transcription factors BosR and BadR [15–26] . In addition to controlling virulence gene expression , BosR also controls expression of genes involved in resistance to reactive oxygen species and affects metal homeostasis , while BadR controls expression of many genes involved in metabolite uptake and utilization [17 , 27 , 28] . Other regulators such as carbon storage regulatory protein A ( CsrA ) appear to exert their effects outside the RpoS/RpoN axis [29] . Much less is known about gene regulation and proteins critical for B . burgdorferi survival while in its tick host [6] . Histidine kinase 1 ( Hk1 ) and response regulatory protein 1 ( Rrp1 ) are highly expressed during the tick phase and appear to work together to regulate expression of genes involved in tick survival [30–32] . Rrp1 is a diguanylate cyclase required for the synthesis of cyclic diguanylate ( c-di-GMP ) , an important second messenger signaling molecule in B . burgdorferi and other bacteria [32–35] The exact mechanisms by which Hk1 is activated and how Rrp1 is regulated are not known . Proteins that have been shown to be important in survival in ticks include outer surface protein A ( OspA ) , which binds to the tick mid gut protein TROSPA [6 , 36] . GuaA and GuaB , two enzymes that contribute to the purine salvage pathway , have also been shown to provide a fitness advantage in the tick host [37] . The glycerol utilization operon ( glpF , glpK , glpD ) encodes proteins that allow the bacterium to utilize glycerol as the carbohydrate source for glycolysis [33 , 38] . This operon is upregulated during all tick life cycle stages , and has been shown to be specifically involved with persistence and survival of the molt , but not early colonization [33 , 35 , 38] . Another protein shown to be essential for infection of the tick host is the manganese transporter BmtA . This transporter is required for B . burgdorferi to colonize and survive in ticks [39] . In this study , we describe the use of transposon insertion sequencing ( Tn-seq ) to identify genes that are critical for B . burgdorferi survival during infection of Ixodes scapularis , the tick vector most commonly associated with Lyme disease transmission in North America [1] . Tn-seq is a high- throughput approach that enables the quantification of the frequency of individual transposon ( Tn ) mutants in a population before and after a selective pressure [40] . Tn-seq has been widely used for in vitro assays of bacterial fitness [41–44] . It has also been used to perform in vivo studies in mice , although in vivo Tn-seq studies are often limited by tight bottlenecks causing stochastic loss of mutants unrelated to the Tn insertion [40–43] . This report represents the first use of a Tn insertional library combined with massively parallel sequencing to identify bacterial genes involved in colonization of an arachnid . Using Tn-seq , we were able to accurately identify a number of B . burgdorferi mutants with impaired fitness for survival in Ixodes ticks . The process is easily scalable though testing additional ticks , which reduces misidentification of mutants that are lost for reasons other than fitness . As opposed to mammalian studies , in which the number of animals is often limiting , we were able to readily screen very large numbers of larval ticks , thereby mitigating bottleneck issues . As part of our studies , we have identified a potential new regulator of B . burgdorferi gene expression , BB0017 , which may contribute to expression of genes involved in tick and mammalian survival .
Mice were bred and maintained in the Tufts University Animal Facility . All experiments were performed following the guidelines of the American Veterinary Medical Association ( AVMA ) as well as the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . All procedures were performed with approval of the Tufts University Institutional Animal Care and Use Committee ( IACUC , Protocol# B2015-159 ) . Euthanasia was performed in accordance with guidelines provided by the AVMA and was approved by the Tufts University IACUC . B . burgdorferi strains were grown in Barbour-Stoenner-Kelley II ( BSK-II ) medium in sealed tubes at 32°C with 1% CO2 . Escherichia coli strains ( Top10 ) for plasmid preparation were grown on Lysogeny broth ( LB ) agar plates or in LB broth at 37oC . E . coli cultures contained either 50 μg/ml spectinomycin or 10 μg/ml gentamicin . The parental strain of the Tn library , the infectious B . burgdorferi strain 5A18NP1 , was used as the wild-type strain in all studies and lacks two plasmids ( lp56 and lp28-4 ) [45] . The following antibiotics were used for selection in cultures of B . burgdorferi when appropriate: kanamycin at 200 μg/ml , gentamicin at 40 μg/ml , and streptomycin at 50 μg/ml . Tn mutants were obtained from the arrayed B . burgdorferi library [45] . All individual Tn mutants used in this study were screened by polymerase chain reaction ( PCR ) at the locus of interest to confirm pure populations as previously described [42 , 43] . In cases where mixed populations were identified ( i . e . two PCR products indicating the presence of both a wild-type and Tn disrupted locus ) , the strain was plated for single colonies in semi-solid agarose overlays . Individual colonies were then selected and re-screened to confirm pure populations . All Tn mutants were subsequently plasmid- typed to identify the loss of any plasmids required for murine or tick infection [46 , 47] . A description of all individual Tn mutants used in this study is available in Table 1 . Other B . burgdorferi strains as well as plasmids used in this study are also described in Table 1 . Infection of ticks with B . burgdorferi from the mutant library was performed using a method previously described [48] . Briefly , before immersion , spirochete cell density was determined by dark field microscopy . Cell suspensions were centrifuged for 10 min at 8 , 000 x g and were resuspended at the desired cell density and in the desired medium . I . scapularis larvae were obtained from the National Tick Research and Education Resource at Oklahoma State University and were maintained in a humid tick incubating chamber at room temperature . Larvae were used within 4 months of emergence . Before immersion , I . scapularis larvae were removed from the chamber and allowed to sit at ambient humidity in an air-conditioned room for 2h . The larvae were then transferred with a small brush into 1 . 5 ml microcentrifuge tubes . B . burgdorferi culture suspension of 108 bacteria in 1 ml was then added to the ticks . For the Tn-seq studies , the suspension consisted of the B . burgdorferi Tn mutant library [42 , 43] . For confirmatory experiments , individual mutants , mixtures of mutants and complemented mutants , or mixtures of mutants and controls were used in the suspension . The tubes were gently vortexed to suspend larvae in the culture and incubated at 32°C for 1 h . Tubes were gently vortexed every 15 min to redistribute ticks in the culture . After incubation at 32°C , tubes were centrifuged at 200 x g for 1 min . The supernatant was removed , and ticks were washed once with phosphate-buffered saline ( PBS ) . Larvae were then transferred from the microcentrifuge tube to a sealed mouse restraining device . The mice were of C57BL/6 background and were all females aged 4-6 weeks old . A mouse was placed into the restrainer containing the larvae , and the larvae were allowed to attach . Mice were removed from the restrainer after 30 min and transferred to cages suspended over water moats . Engorged larvae were collected from the moats 3 to 5 days after placement and transferred into a tick incubation chamber . Ticks were collected as they completed feeding from the animals . Cages were checked daily to collect ticks . Ticks were batched and processed for Tn-seq up to 48 hrs after collection [48] . Briefly , ticks were washed in 3% hydrogen peroxide , 70% ethanol , and finally in PBS . The ticks were allowed to dry before placement in 500 μl of BSK-II medium with kanamycin and gentamicin . To isolate spirochetes from the ticks , the ticks were crushed with a plastic pestle ( Fisherbrand RNase- Free Disposable Pellet Pestles ) . These tick homogenates were inoculated into 5 ml of BSK-II containing kanamycin and gentamicin . The cultures were incubated for two days to allow outgrowth . Following this , the spirochetes were pelleted by centrifugation for 10 min at 8 , 000 x g . Bacterial pellets were washed once in PBS , and then the dry pellet was stored at -80°C until further processing for Tn-seq . Genomic libraries for sequencing were constructed as described previously [42 , 43] . Chromosomal DNA was sheared using the M220 Focused-ultrasonicator ( Covaris ) in microTUBEs with a target peak at 350 bp . The first round of PCR amplification was performed using a modified primer with optimized annealing to the Tn ( pMargent1A , 5'-ggtaccttaggagaccgggg-3' ) [43] . Libraries were multiplexed and pooled for analysis . Sequencing was performed on an Illumina HiSeq 2500 at the Tufts University Core Facility as 50-bp single-end reads , as described previously [42 , 43] . Sequenced reads were clustered by barcode sequence . Data analysis were performed using the Galaxy platform and followed a previously published protocol [43] . We obtained an average of 1 . 2 x 107 reads per barcode , with 1 . 7 x 106 reads per condition for analysis after removal of low quality sequences [43] . Reads were mapped to the B . burgdorferi B31 genome using Bowtie , and a custom script was used to count the number of sequence reads corresponding to each insertion site in the genome . Sequence reads were analyzed “by-site” and “by-gene” . Only Tn mutants that were represented by at least ten sequence reads in both input samples were included in the “by site” analysis . In contrast , the “by-gene” analysis included all sequence reads mapping within a particular gene . Genes represented by less than ten sequence reads in both untreated samples were excluded from the “by-gene” analysis . Tn mutants with zero reads in the output samples were assigned a value of one for the purpose of calculation . The frequency of each Tn mutant in a particular condition was determined by dividing the number of sequence reads corresponding to each Tn mutant by the total number of sequences in the barcode . A frequency ratio was then determined by dividing the frequency of a Tn mutant in the output ( bacteria recovered from the ticks ) sample by its frequency in the input population . For the purposes of prioritizing mutants for follow-up , frequency ratios between 0 . 5 and 2 were considered neutral . A plasmid for complementation of the Tn::bb0164 mutant was generated previously by overlap PCR [43] . A plasmid for directing cis complementation of Tn::bb0412 via allelic exchange was generated from three PCR fragments: intact bb0412 with 995 bp upstream sequence ( F1 ) , PflgB-aadA for antibiotic selection ( F2 ) , and 1022 bp downstream of bb0412 ( F3 ) . Primers were designed with approximately 30 bp on the 5' end for overlap , with the numerically assigned PCR products assembled in order into the final construct . See S1 Table for a list of all primer sequences used in this study . Individual PCR fragments were amplified with AccuPrime Pfx ( ThermoFisher Scientific , MA ) per the manufacturer's instructions . An overlap PCR was performed with equal volumes of the appropriate number of PCR fragments with AccuPrime Pfx reagents per the manufacturer's recommendations . Following PCR , the products were resolved by agarose gel electrophoresis and gel-purified ( Zymoclean Gel DNA Recovery Kit , Zymo Research , Irvine , CA ) for cloning into pCR-Blunt ( ThermoFisher Scientific , Grand Island , NY ) following the manufacturer's protocol . The bb0412 complementation vector was designated pAK412 ( Table 1 ) . Plasmid pMR05 was constructed for trans complementation of the Tn::bb0017 mutant was generated by amplifying a DNA sequence containing the bb0017 open reading frame and the 298-bp upstream region from 5A18NP1 genomic DNA via PCR using the primers bb0017FLAG-F-SacI and bb0017FLAG-R-XbaI ( S1 Table ) . The resulting PCR products as well as pKFSS1 were digested with SacI and XbaI , and the resulting PCR product and pKFSS1 fragments were ligated together using T4 DNA Ligase ( New England Biolabs ) , generating pMR05 ( Table 1 ) . Plasmid pBS01 was constructed to direct allelic exchange at the bb0017 locus , resulting in deletion of the bb0017 open reading frame ( Table 1 ) . Plasmid pBS01 contains 1835 bp of sequence upstream of bb0017 ( amplified from genomic DNA using primers bb0017del1 and bb0017del2 ) , followed by a sequence containing the constitutive PflgB promoter and a streptomycin resistance gene ( aadA , amplified from pKFSS1 using primers bb0017del3 and bb0017del4 ) , followed by 1999 bp of sequence downstream of bb0017 ( amplified from genomic DNA using primers bb0017del5 and bb0017del6 ) . PCR products were purified using the Qiagen PCR purification kit and were subsequently assembled into the pBlueScript cloning vector using the NEBuilder HiFi DNA Assembly Cloning Kit . Plasmid pBS02 was constructed to direct expression of bb0017 in the Δbb0017 mutant . Plasmid pBS02 is identical to pMR05 , except for replacement of the streptomycin resistance cassette by a gentamicin cassette ( Table 1 ) . The streptomycin cassette was excised from pMR05 by restriction digest with AatII and NdeI . The gentamicin cassette had been previously subcloned into the pCR2 . 1cloning vector . The gentamicin resistance gene was excised from this vector using the same restriction enzymes as above , and the resulting fragment was ligated with the digested pMR05 backbone ( Table 1 ) . Plasmid pMH88R was used to constitutively express the glp operon in the Δbb0017 mutant . Plasmid pMH88R was a kind gift of Dr . Frank Yang [33] . All completed plasmids were verified by restriction digest and dideoxy sequencing . Plasmids were introduced into B . burgdorferi by transformation as previously described [50 , 51] . Cis complementation vector pAK412 was transformed into B . burgdorferi Tn::bb0412 , and transformants were designated AK102 ( Table 1 ) . The complemented strain was screened by PCR for allelic exchange using the forward primer of fragment 1 and the reverse primer for the PflgB- aadA cassette for each construct . The bb0017 deletion construct pBS01 was transformed into B . burgdorferi strain 5A18NP1 , generating strain Δbb0017 ( Table 1 ) . In order to complement the bb0017 mutation , plasmid pBS02 was introduced into the Δbb0017 background , generating strain Δbb0017 + bb0017 ( Table 1 ) . The overexpression of the glp operon construct pMH88R was transformed into the Δbb0017 mutant and was used to generate the Δbb0017 + glpFKD strain ( Table 1 ) . Potential transformants were confirmed by PCR with primers designed to detect either a replicating plasmid or a double crossover event , as appropriate , followed by dideoxy sequencing of the PCR product to confirm the expected nucleotide sequence . To evaluate survival in the tick host , individual Tn mutants of interest were combined with their respective complemented strain or wild type bacteria in a 1:1 mixture . Each strain was grown independently . Cell density was determined by microscopy , and 5 x 107 B . burgdorferi were harvested by centrifugation . The pellets from both cultures were then resuspended in the same 1 ml of BSK-II medium to a final overall density of 1 x 108 cells/ml . Ticks were submerged in the cultures as described , placed on mice for feeding , and collected over three days as described above . Each tick was washed successively with 3% hydrogen peroxide , 70% ethanol and PBS , then crushed into 250 μl of BSK-II containing kanamycin and 5 μg/ml amphotericin B . The cultures were allowed to acclimate in liquid medium for a period of 2 h before plating , to allow the bacteria to escape from the crushed tick into the medium . Plating in semi-solid agarose overlay was performed as previously described [50] . These plates were then sealed in plastic bags and placed at 32°C for 10 days . The plates were removed from the incubator and colonies were enumerated . The ratio of the wild-type or complemented strains to the Tn mutant was determined by counting the colonies on the appropriate antibiotic selective plates . A competitive index was calculated for these experiments by dividing the amount of mutant recovered by the amount of complement or WT that was recovered . In the case where no mutant was recovered , its value was set to one for the purposes of calculation of the competitive index . Two independent cultures of the 5A18NP1 and Tn::bb0017 strains were grown to mid-logarithmic phase , washed once in PBS , and resuspended in BSK II medium at a concentration of 6 × 107 bacteria/ml . Cultures were incubated at 32°C with 1% CO2 for 1 h . The bacterial pellet was harvested by centrifugation at 9500 × g for 5 min at 4°C , washed once in ice-cold PBS , and frozen on dry ice . RNA was isolated using the miRNeasy kit ( Qiagen ) . The TURBO DNA-free kit ( Invitrogen ) was used to remove contaminating genomic DNA . Library preparation for RNA-seq analysis was conducted at the Tufts University Genomics Core Facility . Samples were depleted of rRNA using the Gram-Negative Ribo-Zero rRNA Removal Kit ( Illumina ) . Strand-specific libraries were prepared using the TruSeq Stranded mRNA Library Prep Kit ( Illumina ) . Sequencing was performed on an Illumina HiSeq 2500 at the Tufts University Core Facility as 50-bp single-end reads . Data analysis were performed using the Galaxy platform [43] . Sequence reads were aligned to the B . burgdorferi B31 genome using TopHat for Illumina v1 . 4 . 1 with the default settings . Differences in transcript expression were determined using CuffDiff , again using the default settings . Total RNA was extracted from bacterial cells grown to mid-logarithmic growth phase at 32°C using TRIzol ( Invitrogen ) following the manufacturer’s instructions . RNA samples were treated with the TURBO DNA-free kit ( Invitrogen ) to remove contaminating DNA . cDNA was prepared using random hexamers ( Promega ) and the ImProm-II Reverse Transcription System ( Promega ) . Control reactions were performed in the absence of reverse transcriptase to control for the presence of genomic DNA . Sequences of the primers used to determine the differential expression of target genes are listed in S1 Table , and expression levels were normalized against those of the B . burgdorferi housekeeping gene flaB . Quantification of target genes from cDNA was performed using the iTaq Universal SYBR Green Supermix ( BioRad ) [43] . Samples were run in duplicate or triplicate . Analysis of the RT-qPCR data was conducted using the ΔCT method . Data collection was performed using the CFX Connect Real- Time PCR Detection System ( BioRad ) . Strains were grown in BSK-II at 32°C/1% CO2 until cultures reached early stationary phase . A volume corresponding to 1 × 108 bacteria was harvested by centrifugation at 9500 × g for 5 min at 4°C . The bacterial pellet was frozen at -80°C until processing . To lyse the cells , the bacterial pellet was resuspended in approximately 100 μl of 1X NuPAGE buffer ( ThermoFisher ) and boiled for 5 min . A 20 μl volume of each lysate was electrophoresed in 4-15% gradient SDS-PAGE gels ( BioRad ) . Proteins were transferred to a polyvinyldifluoride ( PVDF ) membrane ( Trans-Blot Turbo BioRad ) . Membranes were blocked in 5% milk in Tris-buffered saline containing 0 . 05% Tween-20 ( TTBS ) . Primary antibodies were diluted as follows in TTBS: anti-RpoS ( 1:50 , courtesy of Dr . Frank Yang ) , anti-BosR ( 1:500 , courtesy of Dr . Frank Yang ) , anti-FlaB ( 1:1000 , courtesy of Dr . Xin Li ) , anti-OspC ( 1:10 , 000 , courtesy of Dr . Xin Li ) , anti-OspA ( 1:1000 , Rockland Immunochemicals ) and anti- DbpA ( 1:1000 , Rockland Immunochemicals ) . Appropriate horseradish peroxidase ( HRP ) - conjugated secondary antibodies were used at a 1:10 , 000 dilution in TTBS . Detection was performed using the Luminata Forte substrate ( Millipore ) , followed by exposure to film ( Denville Scientific ) or imaging using a ChemiDoc XRS+ Imager .
In order to use Tn-seq to determine genes involved in B . burgdorferi fitness for survival in ticks , we first needed to decide the number of ticks to include in the experiments . Using immersion feeding , it has previously been shown that approximately 103 B . burgdorferi could be recovered per tick [48 , 52] . However , these experiments examined a single strain of bacteria , and since bacteria were allowed to replicate within the tick before recovery , the actual number of bacteria entering to establish colonization was likely less than 103 . We were also concerned that the bottleneck may be further exacerbated by plasmid loss in the parental strain of the transposon library . It has previously been reported that the mean number of spirochetes per tick with 5A18NP1 lacking both lp28-4 and lp56 plasmids was lower than that in ticks infected with B . burgdorferi harboring all plasmids [53] . The ordered transposon library contains insertions into 45 . 5% of the predicted protein encoding genes [45] . To ensure sufficient coverage of the pooled library of around 4 , 000 mutants we chose to target approximately 150 ticks per experiment . If only half the predicted number of bacteria established infection ( 500 bacteria per tick ) , this approach should still provide approximately 20- fold coverage of the mutants within our library . In each of two independent experiments , approximately 300 larval ticks were immersed in a culture solution containing the entire Tn library and then fed on mice . This resulted in the collection of approximately 160 fed ticks per experiment . Fed ticks were processed and cultured in BSK medium for 2 days . The input culture was also cultured for an additional 2 days to match the tick cultures . Bacteria from both cultures were harvested and sequencing libraries prepared . Reproducibility was high between the two input libraries ( Fig 1A , Pearson coefficient R2=0 . 98 ) . The correlation between the Tn frequencies of the populations recovered from the two groups of tick larvae was also high with a Pearson coefficient R2= 0 . 85 ( Fig 1B ) . A frequency ratio was calculated for each Tn mutant in the library by comparing its frequency in the output library to its frequency in the input library . For analysis , we included only Tn mutants represented by at least 10 sequence reads ( out of a total of approximately 1 . 7 x 106 sequence reads per experiment ) in both replicates of the input libraries . This number was chosen to reduce the risk of stochastic loss after selection in the ticks . Tn mutants that were represented in the input library but had zero reads in the output library were assigned a value of one read in order to be able to calculate a frequency ratio . A frequency ratio less than one indicates that the Tn mutant decreased in frequency after recovery from fed ticks suggesting that the disrupted gene is involved in survival in the larval host . A frequency ratio greater than one indicates that the Tn mutant increased in frequency after larval colonization suggesting that the disruption of the corresponding gene provides a fitness advantage . An overall frequency ratio was also calculated for each gene by aggregating all of the sequence reads mapping to Tn insertions within the same gene ( S1 File ) . A complete list of mutant fitness for larval colonization by gene and by site is provided in S1 File . In order to validate our screen , we began by analyzing genes that have been previously shown to be essential for tick survival , to ensure that these genes had been identified in the screen . Borrelial genes that have been described in the literature as critical to tick survival for which mutants are present in the library include: bb0419 and bb0420 , respectively encoding a response regulator designated Rrp1 and a histidine kinase designated Hk1 [30 , 35]; guaA and guaB , two genes involved in the purine salvage pathway [37]; glpD , encoding a glycerol 3-phosphate dehydrogenase [33 , 38]; and bptA , a surface-expressed lipoprotein [8] . In the Tn-seq experiment , consistent with previously published results , insertional mutants in each of these genes except glpD showed attenuated ability to survive in the tick , with median frequency ratios of <0 . 1 ( Fig 2A ) . GlpD has been shown to be important following the molt from larvae to nymph when carbon sources are less abundant and the organism begins to utilize glycerol , which may explain why Tn::glpD mutants did not exhibit phenotypes in our screen [35] . Further analysis was performed to identify new genes involved in tick survival . A large number of mutants ( N=309 ) had a frequency ratio of less than 0 . 5 compared with the input library . In order to prioritize mutants with the strongest phenotypes for follow-up analysis , we focused on mutants with a fitness ratio of less than 0 . 1 . Mutants with insertions in 102 genes had an average overall frequency ratio below 0 . 1 in both experiments ( S1 File ) . However , this group of 102 mutants included some with insertions into genes that have been shown not to be required for tick colonization ( e . g . ospC ) . While many of this group of 102 genes may be involved in tick survival as it includes many of the genes previously identified as involved in tick colonization , to further reduce the chance of false discovery by the screen , we increased the stringency of our criteria and focused on the subset of 46 genes that had >100 reads in the input library but were completely absent in the processed ticks from both experiments ( Table 2 ) . These genes would be predicted to have the greatest impact on fitness for survival in larval ticks . Of these 46 genes , many have no predicted function and have not been previously characterized ( Fig 3 ) . Approximately 22% of the genes are predicted lipoproteins , while 7% are involved in carbohydrate transport ( Fig 3 ) . To confirm the results of the Tn-seq screen , we chose mutants with insertions in five genes ( bb0017 , bb0164 , bb0412 , bb0050 and bb0051 ) that showed the strongest fitness defects , and that have not previously been reported to be involved in tick survival . Each of these genes was well represented by insertion mutants in the input library and had an overall frequency ratio of less than 0 . 1 following the ingestion of a blood meal by larval ticks . Competition assays were conducted to assess the capability of each individual mutant to survive the larval blood meal ( Fig 2B ) . Three of the transposon mutants were competed against a complemented strain , while the remaining two transposon mutants were competed against the parental strain ( Fig 2B ) . The Tn::bb0017 , Tn::bb0164 , Tn::bb0412 , Tn::bb0050 and Tn::bb0051 mutants were all outcompeted by the parental or respective complemented strains , confirming a role for all five genes in blood meal survival ( Fig 2B ) . We were also able to further confirm this phenotype when competing a bb0017 clean deletion strain against its complemented strain . ( Fig 2B ) . The complemented strain greatly outcompeted the deletion strain confirming the role of bb0017 in surviving the blood meal ( Fig 2B ) . The Tn-seq and competition experiments do not distinguish between the possibilities that 1 ) the identified genes are required for initial entry into the tick during immersion; or 2 ) they are required for surviving the blood meal taken by the tick . The competition experiment that was described previously was modified so that the ticks were not allowed to take a blood meal after immersion in the culture containing the two competing strains , allowing us to separate fitness defects due to uptake from fitness defects due to blood meal survival . The ticks were crushed after two hours of immersion feeding or kept overnight and crushed 24 hours post-immersion . The relative frequencies of the mutant and complemented or parental strains were then determined as before . However , in contrast to the competitive defect exhibited by all five Tn mutants after the blood meal , we were able to recover all Tn mutants after immersion feeding in equal or greater numbers compared to the WT or complemented mutants . However , in the absence of a blood meal , as expected , the numbers of bacteria were greatly reduced and B . burgdorferi was not recovered from all individual ticks . The results of these studies are shown in Table 3 . These data support a role for these five borrelial genes in surviving changes associated with the blood meal . Also , importantly , because several of these genes have identified roles in ROS resistance and hydrogen peroxide was used in washing the ticks , these experiments confirm that the hydrogen peroxide wash did not affect selection of these mutants . To begin to better understand the mechanisms by which the genes identified in the Tn-seq screen contribute to tick-phase survival , we performed further investigation into one of the genes identified as critical for survival of the blood meal: bb0017 . The gene bb0017 was previously identified in a screen for genes that confer resistance to ROS [43] . BB0017 is highly conserved among both B . burgdorferi sensu stricto and other sensu lato strains ( >99% and >94% identity at the amino acid level , respectively ) . BB0017 homologues are also conserved in the relapsing fever strains ( >80% identity ) [54] . In the B . burgdorferi strain B31 , bb0017 is annotated an integral membrane protein of the YitT family . BB0017 contains four predicted transmembrane domains as well as a C-terminal soluble domain and contains the conserved domain of unknown function DUF2179 ( S1A & S1B Fig ) [55] . A structure-based similarity search using Phyre2 suggested that the C-terminal domain of BB0017 is structurally similar to PII and PII-like proteins , despite low overall sequence identity ( <27% identity ) [56] . No high confidence predictions were made for the N-terminal domain of BB0017 . PII proteins are a broadly conserved class of signal transduction proteins found in bacteria , archaea , and plants and are generally small cytoplasmic proteins involved in nitrogen metabolism [57–59] . PII proteins generally function as trimers and control the activity of their regulatory targets through direct protein-protein interactions in response to both post-translational modifications ( such as uridylylation ) and ligand binding ( including ADP , ATP , and 2-oxoglutarate ) . The long , flexible T- loop mediates interactions with regulatory targets , while a conserved motif in the shorter B-loop is involved in ligand binding ( S1C Fig , GlnKEc ) . More recently , several PII-like families of proteins have been identified in bacteria , including a family of proteins in Gram-positive bacteria that bind cyclic diadenylate monophosphate ( c-di-AMP ) as well as a broadly conserved family of proteins ( CutA ) that confer copper tolerance in Escherichia coli and bind acetylcholinesterase in mammals [60–66] . While the PII and PII-like proteins share a common ferredoxin-like fold , the lengths of the T and B loops differ significantly between the different protein families . In the case of the PII-like c-di-AMP binding proteins , the lengths of the B and T loops are reversed relative to the PII proteins and are referred to as the B´ and T´ loops ( PstASa , S1C Fig ) . Structural data suggests that the functions of the B´ and T´ loops are also reversed relative to the PII proteins , with the short T´ loop being involved in ligand binding and the long flexible B´ loop possibly involved in effector binding [61 , 62 , 64] . In the case of the copper tolerance protein CutA1 , both the B and T loops are truncated , and the same is true for BB0017 . Interestingly , BB0017 appears to lack conserved residues involved in ligand binding by both the PII and PII-like protein families . The presence of the N-terminal transmembrane domain also distinguishes BB0017 from the PII and PII-like protein families and suggests that membrane localization may be important for BB0017 function . Because BB0017 contains a putative signal transduction domain , we hypothesized that the Tn::bb0017 mutant would exhibit global differences in gene expression compared to the parental strain . Total RNA was isolated from the parental and Tn::bb0017 strains , and RNA sequencing ( RNA-seq ) was used to compare the transcriptomes of both strains . We identified 16 genes that were significantly downregulated more than twofold in the Tn::bb0017 mutant compared to the parental strain and 25 genes that were significantly upregulated more than twofold ( S2 Table and Fig 4 ) . It is important to note that bb0017 does not appear in S2 Table . While expression of bb0017 was significantly different between the Tn::bb0017 and parental strains , the difference was less than twofold , and bb0017 expression was actually higher in the Tn::bb0017 mutant compared to the parental strain . Sequence coverage maps confirm that transcription in the Tn::bb0017 mutant is abrogated downstream of the Tn insertion as expected ( S2 Fig ) . However , increased numbers of sequence reads mapped in the 5’ portion of bb0017 , likely due to transcription from the strong PflaB promoter contained within the Tn ( S2 Fig ) . It is unclear whether there is translation of the 5’ portion of bb0017 in the Tn::bb0017 mutant , resulting in a truncated protein , but if a truncated protein is produced in the Tn::bb0017 mutant , these results could suggest that the C-terminal portion of BB0017 is the critical portion for survival in the tick . Strikingly , the putative bb0017 regulon overlaps significantly with that of RpoS , a key regulator of virulence gene expression in B . burgdorferi [15 , 16 , 19 , 20] . RpoS is directly responsible for upregulating a number of genes required for survival in the mammalian host , including dbpA , dbpB , ospC , and bbk32 and repressing expression of genes important for tick survival such as glpD [15 , 19 , 20 , 38 , 67] . The dbpA , dbpB , ospC , and bbk32 genes are all upregulated in the Tn::bb0017 mutant ( S2 Table , Fig 4 ) . Several genes known to be subject to RpoS-mediated repression , including genes located within a glycerol utilization operon important for tick infectivity , are downregulated in the Tn::bb0017 mutant ( bb0240-bb0243 , S2 Table ) [38]; however , other regulators such as c-di-GMP may also affect expression of these genes . To confirm the results of the RNA-seq screen , we generated a mutant lacking the entire bb0017 open reading frame . A survey of available B . burgdorferi genome sequences revealed two different annotated start sites ( S1 Fig ) . We chose to delete the region encompassing the first start site , which includes a putative 71 bp small RNA ( SR0011 ) in the bb0016-bb0017 intergenic region [68] ( S1 Fig ) . We restored bb0017 expression including the upstream SR11 intergenic region under the control of the native promoter from a replicating plasmid in the Δbb0017 mutant , and confirmed expression by qRT-PCR ( S3 Fig ) . We performed qRT-PCR on the Δbb0017 mutant as well as the complemented strain to validate the results of the RNA-seq using the transposon insertion strain . We selected a subset of differentially regulated genes from the RNA-seq , as well as some representative regulatory proteins of B . burgdorferi that showed no change in expression . bosR and rpoS levels were not significantly different in the RNA-seq and this phenotype was reproduced in the deletion strain as well as the complement by qRT-PCR ( Fig 5 ) . We then confirmed six genes , ospC , dbpA , glpD , bba37 , bba25 , and bbk32 that were differentially expressed by RNA-seq in the transposon mutant strain . Each showed a similar pattern of expression in the clean deletion strain with recovery in the complemented mutant strain , with the exception of glpD ( Fig 5 ) . Transcription of glpD was decreased in the transposon mutant and its complement as well as the deletion strain and its complement in comparison to the wild type making it unlikely that this difference was due to secondary site mutations or polar effects as each of the mutants and complements were created from separate isolations from the parental strain . Of note is that all the changes are small ( fourfold ) compared to the other genes tested by qRT-PCR . Given the lack of involvement of glpD involvement in tick survival as assayed by the Tn-seq and its established role at a different stage in tick survival , it is likely that the change is not physiologically relevant . To ensure that we were not missing a role for GlpD in mediating effects of the Δbb0017 mutant , we created a strain that overexpresses the entire glp operon , including glpFKD , in the Δbb017 deletion strain . This construct has previously been used to successfully overexpress GlpD [33] . We confirmed that glpD was successfully transcriptionally over-expressed by qRT-PCR ( Fig 5 ) . Using this strain , we then performed a competition experiment between the glp operon overexpressing strain and the Δbb0017 strain . We were not able to recover either strain from this experiment following selective plating from six collected fed larvae . This indicates that increasing the ability for glycerol utilization is not sufficient to rescue the Δbb0017 mutant and the defect in tick colonization is unlikely to result strictly from decreased expression of glpD . We performed immunoblots for OspC and DbpA in the Tn::bb0017 and Δbb0017 mutants . Levels of both DbpA and OspC were elevated in the Tn::bb0017 and Δbb0017 mutants , confirming the results of the RNA-seq screen ( Fig 6A ) . Restoration of bb0017 expression from a replicating plasmid ( which also contains SR0011 ) in both mutants decreased DbpA and OspC levels to those of the parental strain ( Fig 6A ) . The fact that both the Tn::bb0017 mutant ( in which SR0011 remains intact ) and the Δbb0017 mutant ( in which SR0011 is disrupted ) exhibit increased lipoprotein expression , suggests that bb0017 is required for the phenotype , although these results to do not exclude the possibility that SR0011 may also be involved . Expression of OspA , a surface lipoprotein required for infectivity in the tick , was not affected by the absence of bb0017 under the conditions tested . The expression of dbpA and ospC requires the alternative sigma factor RpoS [15 , 20] . The regulation of rpoS in turn involves a second alternative sigma factor RpoN ( σ54 ) , the enhancer binding protein Rrp2 , the Borrelia oxidative stress response regulator BosR , and the small regulatory RNA DsrA [15 , 16 , 24 , 25 , 69 , 70] . We hypothesized that BB0017 mediates the repression of ospC and dbpA indirectly by affecting the expression of an upstream regulator . We therefore investigated RpoS levels in the Tn::bb0017 and Δbb0017 mutants . RpoS levels were increased in both mutants , and restoration of bb0017 expression resulted in decreased RpoS levels ( Fig 6B ) . To understand the mechanism by which BB0017 affects RpoS expression , we next investigated production of BosR , a positive regulator of RpoS [24 , 70] . As was the case for RpoS , BosR levels were increased in the Tn::bb0017 and Δbb0017 mutants , and complementation of bb0017 restored levels to those similar to the parental strain ( Fig 6B ) .
In this paper , we report the use of massively parallel , next generation sequencing technology to identify genes important in survival in the larval tick host . This study represents the most complete survey of B . burgdorferi genes that are required for tick survival performed to date and greatly increases our understanding of this critical phase of the B . burgdorferi life cycle . We have identified many genes that have not previously been associated with tick survival , confirmed the involvement of a subset in tick survival , and began to characterize a mechanism of action for one of the genes , bb0017 . Notably , because we were able to quickly and inexpensively screen large numbers of ticks , we were able to minimize bottleneck issues that have arisen in other animal studies , and our results showed a high level of experimental reproducibility . The robustness of the technique is exemplified by our ability to identify genes known to be required for tick-phase survival and our ability to validate the phenotypes of all five mutants we selected for further analysis . There are several caveats in the interpretation of the Tn-seq screen data . First , the transposon library is not saturated and does not contain insertions into all non-essential genes . Mutants for several genes known to be important in tick colonization are not included in the library . Next , the transposon library we used for this study was generated in the 5A18NP1 background , which is missing two plasmids , lp56 and lp28-4 . It is possible that the loss of the genes on these plasmids affects the requirement for certain genes or that the regulatory patterns are altered in their absence . Finally , the mechanism we used to infect the tick larvae , immersion feeding , is artificial and may lead to identification of genes that are not involved in natural transmission or , more likely , miss genes that are involved . A more general caution about screening techniques such as Tn-seq is that the false discovery rate is dependent upon the stringency of the analysis used . We analyzed the data in two tiers . Using our most stringent criteria of no mutants isolated in either replicate , we did not detect any mis-identification in the subset of genes that were confirmed by additional experimental testing . Using slightly less stringent criteria of a 90% decrease in fitness , we already noted some false identifications of genes that have been previously shown to not be involved in tick survival including ospC . As with any screen , the goal is to enrich the identification of genes that are actually involved in a process while minimizing false identifications , but regardless of the stringency of the criteria , each of the genes will still need to be confirmed by additional testing . There was quite a bit of variability in frequency between genes that were not completely absent . This occurs because of stochastic loss of mutants due to bottleneck issues that can result in differences in the recovered mutants . One way to minimize this variability is to perform more experiments , which in our case , would mean adding more ticks for each replicate . By averaging results over more experiments , stochastic variability will decrease and we would have increased ability to identify genes with partial fitness impacts . At the numbers of ticks we used , the greatest confidence is for an extreme phenotype . Understanding these caveats , we identified 46 genes whose disruption resulted in complete loss of Tn mutants from the population following colonization of larval ticks . Of these 46 genes , almost all have not previously been reported to be involved in survival in the tick , and thus the current study represents a significant advancement in our understanding of the genetic factors required for B . burgdorferi survival in the tick . Of note , many of the tick-phase genes we identified encode membrane-localized lipoproteins , and a significant portion ( 14 genes , including the five we selected for follow-up analysis ) , have been previously identified as important for resistance to reactive oxygen and nitrogen species ( Table 2 ) [43] . We confirmed the phenotype for five of the novel tick-phase genes ( bb0017 , bb0164 , bb0412 , bb0050 , and bb0051 ) by individual competition assays for survival in the tick following a blood meal ( Fig 2B ) . Prior to our study , relatively little was known regarding the functions of these gene products , other than their predicted role in ROS resistance . All five gene products are predicted to be membrane-localized , and BB0164 has previously been shown to be involved in controlling intracellular manganese homeostasis [43] . To better understand whether these genes aid 1 ) the entry of B . burgdorferi into tick larvae during artificial infection , or 2 ) bacterial adaptation as the tick takes a blood meal , additional competition studies were carried out following the culture immersion step , but before the blood meal . All of the mutants tested survived as well as ( or better than ) the controls in the competition assays , suggesting that these genes are involved with survival of the blood meal and not with entry into the tick . That survival of the blood meal poses the larger barrier is not surprising . The blood infusion that the B . burgdorferi encounters in the midgut of the tick during feeding creates a rapidly changing environment for the spirochete . During the blood meal , there are changes in pH and temperature and exposure to reactive oxygen species ( ROS ) , natural antibodies , and components of complement that can mediate spirochete killing [71–78] . Our results suggest that the ability to resist oxidative stress is likely critically important for survival in the tick host . We performed further investigations to better understand the mechanisms by which one of the genes identified in our screen , bb0017 , contributes to survival in the tick . Our in silico analysis suggested that BB0017 is part of a larger family of PII and PII-like proteins . However , there are some notable differences that distinguish BB0017 from these protein families , including differences in the lengths of two key loop regions ( B/B’ and T/T’ ) and absence of key conserved residues involved in ligand binding . Thus , if BB0017 does bind ligands such as copper or c-di-AMP as has been previously shown for PII and PII-like proteins , it does so via a unique mechanism . c-di-AMP is produced in B . burgdorferi , although its potential function as a second messenger in this organism remains unclear [79 , 80] . It is certainly possible that the true ligand for BB0017 is a different molecule as there are significant differences between BB0017 structure and the structure of other PII proteins . Given the downstream effects of BB0017 , it is tempting to speculate that it may bind c-di-GMP , which plays a critical role in B . burgdorferi gene regulation , however identification of binding of other molecules by BB0017 requires further experimentation . RNA-seq analysis revealed that interruption of bb0017 by the Tn insertion results in significantly higher levels of transcription of the genes encoding DbpA and OspC; these results were corroborated by qRT-PCR as well as by immunoblot analyses using strains that had a complete deletion of bb0017 ( S2 Table , Fig 5 and Fig 6A ) . This regulatory effect in the mutant appears to be mediated by increased levels of BosR and RpoS suggesting that BB0017 acts as a potential negative regulator of these important pathways . Repression of genes highly expressed during mammalian infection would be consistent with a role for BB0017 in tick colonization . Previous studies have shown that expression of B . burgdorferi genes that are required for one host may result in a fitness defect in colonization of the other host [69 , 73 , 76 , 77 , 81–83] . The effects of deletion of bb0017 on RpoS are also likely to affect expression of genes in the glp operon as seen by the RNA-seq studies ( S2 Table ) . However , altered expression of the glp operon does not appear to account for the survival defect of the bb0017 mutant in our Tn-seq experiments as overexpression of the glp operon was not sufficient to restore the ability to survive the blood meal in a Δbb0017 background . This is consistent with the fact that we did not observe a fitness defect for the Tn::glpD mutant and that prior studies have shown that the glp genes are required at a time point later in the tick cycle than was evaluated in our study [33 , 38] . The elevated lipoprotein expression profile observed in the bb0017 mutant is strikingly similar to the phenotype of a mutant lacking the BmtA manganese ( Mn ) transporter . The bmtA mutant exhibits decreased intracellular Mn concentrations , which was shown to result in increased levels of ospC expression [84] . In the case of the bmtA mutant , the increased ospC expression is due to an increase in BosR protein levels at the post-transcriptional level , leading to increased transcriptional activation of RpoS [84] . The post-transcriptional regulation of BosR has also been observed in conditions where CO2 is limiting [85] . Our RNA-seq analysis suggests that bosR and rpoS transcript levels are similar in the Tn::bb0017 mutant and parental strains , despite the increase we observe in protein levels ( Figs 5 and 6 ) suggesting that BB0017 affects BosR at the post-transcriptional level . In the bb0017 mutant , RpoS also appears to be upregulated at the post-transcriptional level , suggesting that some mechanism other than direct transcriptional activation by BosR is responsible for increased RpoS levels . There is also precedent for post-transcriptional regulation of RpoS , both in B . burgdorferi and in other bacteria [69] . The reciprocal expression of two sets of genes required for survival in the tick or mammalian hosts in response to a variety of environmental signals is paradigmatic to borrelial pathogenesis . However , the mechanisms by which these external stimuli are sensed remain to be fully characterized [69 , 73 , 76 , 77 , 81–83] . Given that BB0017 is predicted to be a membrane-localized signal transduction protein , we hypothesize that this protein may sense changes in the environment to regulate downstream effectors accordingly . The nature of the external signal , if any , sensed by BB0017 remains unclear , although it is likely not Mn . We previously showed that Mn levels are similar in the Tn::bb0017 mutant compared to the parental strain [43] . Given that PII and PII-like proteins generally affect downstream targets at the post-transcriptional level via direct protein- protein interactions , we predict that the BB0017 regulon may be larger than the list of genes identified in the RNA-seq analysis . In conclusion , we have found that Tn-seq is a powerful tool in identifying B . burgdorferi genes important for fitness in surviving the blood meal . We have identified a large number of previously uncharacterized genes involved in the survival of the bacterium in its tick host . These results may provide important new avenues for exploration and understanding how the bacterium adapts to its different hosts . To this end we have further investigated the role of one of the genes identified , bb0017 . We propose that BB0017 is a potential global regulator in B . burgdorferi that affects resistance to oxidative stress , survival in the arthropod host , and expression of key virulence determinants . As several of the other genes identified as important for survival of the bacteria during the early stages of tick larval infection also have been identified in prior screens for genes of ROS resistance , our results suggest the importance of ROS resistance in the initial colonization and persistence during the acquisition of the blood meal . Future Tn-seq screens can be tailored to identify genes required for survival during other parts of the bacterial lifecycle within the tick host . This approach will allow investigators to map the network of adaptations used by the bacteria to complete its life cycle .
|
Borrelia burgdorferi , the causative agent of Lyme disease , must adjust to environmental changes as it moves between its tick and vertebrate hosts . We performed a screen of a B . burgdorferi transposon library using massively parallel sequencing ( Tn-seq ) to identify fitness defects involved in survival in its tick host . This screen accurately identified genes known to cause decreased fitness for tick survival and identified new genes involved in B . burgdorferi survival in ticks . All of the genes tested individually confirmed the Tn-seq results . One of the genes identified encodes a protein whose function was previously unknown that appears to be involved in regulating expression of proteins known to be involved in environmental adaptation . Tn-seq is a powerful tool for understanding vector-pathogen interactions and may reveal new opportunities for interrupting the infectious cycle of vector-borne diseases .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"lipoprotein",
"metabolism",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"microbiology",
"developmental",
"biology",
"molecular",
"biology",
"techniques",
"bacteria",
"lipoprotein",
"structure",
"bacterial",
"pathogens",
"research",
"and",
"analysis",
"methods",
"borrelia",
"burgdorferi",
"artificial",
"gene",
"amplification",
"and",
"extension",
"proteins",
"medical",
"microbiology",
"gene",
"expression",
"microbial",
"pathogens",
"life",
"cycles",
"lipoproteins",
"borrelia",
"molecular",
"biology",
"biochemistry",
"blood",
"anatomy",
"polymerase",
"chain",
"reaction",
"genetic",
"screens",
"physiology",
"gene",
"identification",
"and",
"analysis",
"genetics",
"biology",
"and",
"life",
"sciences",
"larvae",
"organisms"
] |
2019
|
Genome-wide screen identifies novel genes required for Borrelia burgdorferi survival in its Ixodes tick vector
|
Taenia solium taeniosis/cysticercosis is a parasitic infection occurring in many developing countries . Data on the status of human infections in Zambia is largely lacking . We conducted a community-based study in Eastern Zambia to determine the prevalence of human taeniosis and cysticercosis in a rural community . Stool and serum samples were collected from willing participants . Geographical references of the participants' households were determined and household questionnaires administered . Taeniosis was diagnosed in stool samples by coprology and by the polyclonal antibody-based copro-antigen enzyme-linked immunosorbent assay ( copro-Ag ELISA ) , while cysticercosis was diagnosed in serum by the B158/B60 monoclonal antibody-based antigen ELISA ( sero-Ag ELISA ) . Identification of the collected tapeworm after niclosamide treatment and purgation was done using polymerase chain reaction-restriction fragment length polymorphism ( PCR-RFLP ) . A total of 255 households from 20 villages participated in the study , 718 stool and 708 serum samples were collected and examined . Forty-five faecal samples ( 6 . 3% ) were found positive for taeniosis on copro-Ag ELISA while circulating cysticercus antigen was detected in 5 . 8% ( 41/708 ) individuals . The tapeworm recovered from one of the cases was confirmed to be T . solium on PCR-RFLP . Seropositivity ( cysticercosis ) was significantly positively related to age ( p = 0 . 00 ) and to copro-Ag positivity ( taeniosis ) ( p = 0 . 03 ) but not to gender . Change point analysis revealed that the frequency of cysticercus antigens increased significantly in individuals above the age of 30 . Copro-Ag positivity was not related to age or gender . The following risk factors were noted to be present in the study community: free-range pig husbandry system and poor sanitation with 47 . 8% of the households visited lacking latrines . This study has recorded high taeniosis and cysticercosis prevalences and identified the need for further studies on transmission dynamics and impact of the disease on the local people .
Taenia solium taeniosis/cysticercosis is a neglected parasitic zoonosis , affecting mostly developing countries [1] . Its occurrence is strongly associated with poverty , poor hygiene and sanitation , free-range pig management and lack of meat inspection [2]–[4] . Adult intestinal tapeworm infection ( taeniosis ) in man , who is the sole natural definitive host [5] , is acquired by eating undercooked infected pork . Infective eggs are shed ( with proglottids ) via the stool and contaminate the environment . Pigs are the intermediate host and are infected by ingestion of these infective eggs ( or proglottids ) , which develop into cysticerci ( porcine cysticercosis ) . Humans can act as intermediate hosts as well; cysts can develop subcutaneously , intramuscularly , but often in the central nervous system causing neurocysticercosis ( NCC ) . NCC has been described as the most frequently reported helminthic infection of the central nervous system [6] and is a major cause of acquired epilepsy in cysticercosis endemic regions associated with considerable morbidity [7] . In the last decade , many studies have been carried out in sub-Saharan Africa to determine the presence/absence of T . solium . Until now , most studies have been carried out on porcine cysticercosis reporting endemicity in countries like Tanzania , Zambia , Mozambique and Kenya [8]–[12] . Prevalence of human cysticercosis , which has been less studied , ranges from 7 . 4% in South Africa to 20 . 5% in Mozambique ( both based on specific antibody detection ) and 21 . 6% in the Democratic Republic of Congo ( based on circulating antigen detection ) [13]–[16] . In studies conducted in Kenya , taeniosis prevalences were estimated from 4 to 10% [17] . These data emphasize the need for more studies in humans to gather information on the epidemiology of the parasite and to estimate the burden on affected communities . In Zambia , pig keeping and pork consumption have increased significantly during the past decade with Eastern and Southern provinces accounting for a greater part of this increase [4] . Pigs are mostly kept by smallholder producers , under free-range management . Several studies carried out in Zambia have indicated high prevalences of porcine cysticercosis . A study at a slaughter slab in Lusaka receiving village pigs , indicated a prevalence of up to 64 . 2% [18] while field studies in Southern and Eastern province estimated sero-prevalences between 10 . 8–20 . 8% and 9 . 3–23 . 3% respectively [19] , [20] . These studies clearly showed that T . solium was present in rural areas of Zambia . However , except for preliminary unpublished data , no information was available on human cysticercosis/taeniosis . The main objective of the current study was to determine the prevalence of human taeniosis and cysticercosis in a rural community of Petauke district in the Eastern province of Zambia .
Ethical clearance ( approval ) for both human and animal parts of the study was obtained from the University of Zambia Biomedical Research Ethics Committee ( IRB0001131 ) . For the human part of the study , further approval was sought from the Ministry of Health of Zambia and also from the local District health authorities . A meeting was held with the community leaders ( headmen ) to explain the purpose of the study and request their permission to conduct the study in their area . Finally consent was also sought from the individual subjects to take part in the study . Subjects were not forced to participate and participation was requested of individuals of all ages after written informed consent . For individuals below the age of 16 , permission was sought from their parents or guardians by way of written informed consent . All participants found positive for taeniosis and other helminths were provided with treatment , namely niclosamide and mebendazole respectively . Those positive for cysticercosis were referred to the District hospital for follow-up and the recommended standard of care provided to them if required . Collection of reference samples from pigs at a local abattoir was carried out by professional veterinarians adhering to the Zambian regulations and guidelines on animal husbandry . The study was conducted in a rural community ( Kakwiya ) in Petauke district of the Eastern province of Zambia ( Figure 1 ) . The community is serviced by a Rural Health Centre ( RHC ) whose catchment population is 11 , 344 ( Clinic headcount records ) . The people in this community practice subsistence farming growing mostly crops like maize and groundnuts primarily for home consumption and cotton and bananas grown for household income generation . They also keep animals , mostly pigs with a few keeping cattle , goats and chickens . A preliminary visit to the area indicated that , as reported by Sikasunge et al . [21] , there was a high number of free roaming pigs and reports of cysts observed in pigs slaughtered in backyards . A community-based cross sectional survey was conducted in the dry season between July and August 2009 . The Kakwiya community was selected because it is a pig keeping community without any active ongoing sanitation programmes and cysticerci were observed in slaughtered pigs . The willingness of the community to participate in the study and the RHC to collaborate was also taken into account together with the availability of an adequate working space in the clinic and staff for the collection of samples . Only villages ( n = 20 ) within a radius of 7 km from Kakwiya RHC were selected . The selected villages were visited and all persons invited to participate in the study . Each willing participant was , after written informed consent , given two plastic sample bottles and requested to submit a stool and a urine sample at the RHC . Upon submission of these samples , a blood sample was also taken by qualified health personnel . A questionnaire targeting household characteristics ( such as number of inhabitants in a household , main household income , highest level of education attained , main source of drinking water ) , presence of household level risk factors ( such as keeping of pigs and how they are kept , backyard slaughter of pigs , inspection of slaughtered pigs , consumption of pork by at least one member of the household , presence of a pit latrine , consumption or resell of clearly infected pork ) and knowledge of the parasite ( such as observation of tapeworm in human faeces , how people acquired a tapeworm , observation of cysts in pork meat , what the cysts were and how pigs acquired them ) was administered to each participating household . At the same time geographic co-ordinates of each participating household were obtained using a Global Positioning System ( GPS ) receiver ( eTrex Legend® Cx , Garmin ) . Submitted stool samples were divided into two aliquots; one placed in 10% formalin and the other in 70% ethanol and these were stored at 4°C until use . Urine samples were aliquoted in 1 . 8 ml vials and stored at −20°C . Analyses and results for the urine specimens were discussed in our earlier report [22] . About 5 ml of blood were collected into sterile plain blood collecting tubes and allowed to clot . To obtain the maximum amount of serum , the blood tubes were allowed to stand at 4°C overnight and then centrifuged at 3000 g for 15 minutes . The supernatant ( serum ) was aliquoted into 1 . 8 ml vials and stored at −20°C until use . All the samples were transported to Lusaka where they were stored at −20°C until analysis . Presence of helminth ova in stool was examined microscopically using the formalin-ether concentration technique [23] . Presence of a taeniid egg on a slide was recorded as being positive for taeniosis . The presence of other helminth eggs was also noted during the examination . An in house copro-antigen detection ELISA ( copro-Ag ELISA ) as described by Allan et al . [24] , with slight modifications , was performed on the stool samples . Briefly , the stool samples stored in 10% formalin were processed by mixing an equal amount of Phosphate Buffered Saline ( PBS ) and stool sample . This was allowed to soak for one hour with intermediate shaking and centrifuged at 2000 g for 30 minutes . The supernatant was then used in the ELISA . The assay involved coating polystyrene ELISA plates ( Nunc® Maxisorp ) with the capturing hyper immune rabbit anti-Taenia IgG polyclonal antibody diluted at 2 . 5 µg/ml in carbonate-bicarbonate buffer ( 0 . 06 M , pH 9 . 6 ) . After coating , the plates were incubated for 1 hour at 37°C , washed once with PBS in 0 . 05% Tween 20 ( PBS-T20 ) and all wells blocked by adding blocking buffer ( PBS-T20+ 2% New Born Calf Serum ) . After incubating at 37°C for 1 hour and without washing , 100 µl of the stool supernatant was added and plates were incubated for 1 hour at 37°C followed by washing five times with PBS-T20 . A biotinylated hyper immune rabbit IgG polyclonal antibody diluted at 2 . 5 µg/ml in blocking buffer was used as the detector antibody . One hundred microlitre was added and the plates were incubated for 1 hour at 37°C followed by washing 5 times . One hundred microlitre of Streptavidin-horseradish peroxidase ( Jackson Immunoresearch Lab , Inc . ) diluted at 1/10 , 000 in blocking buffer was added as conjugate . After 1 hour incubation at 37°C and washing 5 times , 100 µl of ortho phenylenediamine ( OPD ) substrate , prepared by dissolving one tablet in 6 ml of distilled water and adding 2 . 5 µl of hydrogen peroxide , was added . The plates were incubated in the dark for 15 minutes at room temperature before stopping the reaction by adding 50 µl of sulphuric acid ( 4 N ) to each well . The plates were read using an automated spectrophotometer at 490 nm with a reference of 655 nm . To determine the test result , the optical density ( OD ) of each stool sample was compared with the mean of a series of 8 reference Taenia negative stool samples plus 3 standard deviations ( cutoff ) . Presence of circulating cysticercus antigens was measured by the monoclonal antibody based B158/B60 Ag-ELISA ( sero-Ag ELISA ) [25] , [26] . Sera from two known highly positive pigs ( obtained from a local pig market and confirmed by dissection ) were used as positive controls . The OD of each serum sample was compared with a sample of negative serum samples ( N = 8 ) at a probability level of P = 0 . 001 to determine the result in the test [26] . Differentiation of the Taenia spp . was done using molecular methods . Taeniosis positive individuals were treated with niclosamide ( 2 g single dose ) followed by a purgative ( Magnesium sulphate , 30 g ) . The collected tapeworm segments were stored in 70% ethanol until use . DNA was extracted from the parasitic material using the Boom extraction method slightly modified as described by Rodriguez-Hidalgo et al . [27] and PCR used to amplify the mitochondrial 12 s rDNA gene fragment . Restriction fragment length polymorphism ( RFLP ) was then used to differentiate the Taenia spp . [27] . All collected data was entered into an Excel ( Microsoft Office Excel 2007® ) spreadsheet and analyses were conducted in Stata 10 ( http://www . stata . com ) . Chi square test was used to check for differences between disease positivity and gender . Uni- and multivariate logistic regressions were used to investigate the relationship between taeniosis and cysticercosis positivity and individual gender and age . The age variable was first used as a continuous variable and then categorized into 10 categories of 10 years each , in order to identify changes in positivity frequencies as a function of the age of individuals . A change point analysis was used to simplify the observed relations into antigen patterns as a function of age [28] , [29] . The level of significance was set at p<0 . 05 for all statistical analyses . The geo-reference data collected was used for spatial analysis using ArcView Gis 3 . 2 ( http://www . esri . com ) . Analysis of the possibility of geographical clustering of households with latrines or those that kept pigs and also cases of taeniosis and cysticercosis was done by means of the risk-adjusted nearest neighbour hierarchical spatial clustering ( Rnnh ) using Crimestat® III [30] . Given the limited number of individuals infected with taeniosis and cysticercosis , the minimum number of cases per cluster was set at 3 while the minimum number of households with a latrine or that kept pigs was set at 20 . Monte Carlo simulations were run in this software to determine the significance of the clusters . Significance level of a cluster in the simulation was set at 95% and a cluster was determined significant if the density of the points was higher than that obtained at the 95th percentile after 1000 simulations .
A total of 720 willing participants from 20 villages belonging to 255 households participated in the study . Of these , 428 ( 59 . 4% ) were females and 292 ( 40 . 6% ) were males and the age ranged from 1 to 96 years with a median age of 12 years . The age distribution , with a majority of the younger age group , was typical of rural areas in developing countries [31] . The number of people living in a household ranged from 1 to 13 with a median of 7 . Seven hundred and eighteen of the participants gave a stool sample and 708 gave a blood sample . At least one participant from each participating household gave a sample depending on the willingness of the household members . The number of individuals sampled from each household ranged from 1 to 11 . Some household characteristics recorded from the questionnaire administered to the 255 households included; 32 . 6% kept pigs with 99 . 6% of these rearing on free-range , 47 . 8% of the households did not have latrines ( Figure 2 ) and 94 . 5% of the households had at least one individual who consumed pork . Three clusters each of households with latrines ( 14 . 13881S , 31 . 19501E , density = 748 . 97; 14 . 14338S , 31 . 20369E , density = 151 . 95 and 14 . 09718S , 31 . 17940E , density = 117 . 15; 95th percentile density = 0 . 02 ) and those that kept pigs ( 14 . 13891S , 31 . 19493N , density = 299 . 89; 14 . 14390S , 31 . 20374E , density = 134 . 33 and 14 . 09773S , 31 . 17961E , density = 86 . 79; 95th percentile density = 0 . 01 ) were identified in the study area ( Figure 2 ) . About 44% of the households reported to have slaughtered a pig in their backyards . None of them had the meat inspected before either home consumption or resell to members of the community . Pork was reported to be consumed in a variety of ways including boiling , frying and roasting . The data obtained in the questionnaire on risk factors is described in more detail in another report ( Mwape et al . , submitted article ) . The results for both the coproscopic examination and copro-Ag ELISA are shown in Table 1 . Two ( 0 . 3% ) individuals were positive for taeniosis on coproscopic examination while copro-Ag ELISA detected 45 ( 6 . 3% ) positives . The two coproscopic positives were also positive on copro-Ag ELISA . Figure 3 shows the copro-Ag ELISA results in function of 10 age groups of 10 years each . The highest prevalence was determined in the 80–89 years age group , though this was not significantly different from the other age groups . A univariate logistic regression analysis did not indicate any relationship between copro-Ag ELISA positivity and sex ( p = 0 . 548 ) or age ( p = 0 . 311 ) . This finding was the same for the multivariate analysis with p values of 0 . 139 and 0 . 645 for sex and age respectively . One cluster of taeniosis cases ( 14 . 13868S , 31 . 19509E , density 116 . 55; 95th percentile density = 49 . 33 ) was identified in the study community ( Figure 4 ) . At household level , the number of positives per household ranged from 0 to 3 . All taeniosis positive individuals were treated with 2 g niclosamide per os and given a purgative ( Magnesium sulphate ) two hours later . One tapeworm was collected and confirmed to be T . solium by PCR-RFLP . The results for the sero-Ag ELISA are shown in Figure 5 in function of 10 age groups of 10 years each . Circulating cysticercus antigens were detected in 41 ( n = 708 ) participants giving an apparent prevalence of 5 . 8% ( 95% CI , 4 . 1–7 . 5 ) . Uni- and multivariate logistic regression analysis revealed a very strong relationship between sero-Ag positivity and age ( p<0 . 001 ) . Figure 6 indicates that the prevalence of cysticercosis is initially low and a change point analysis indicated a significant increase in positivity frequencies at 30 years of age . The logistic regression model indicated that the proportion of sero-Ag ELISA positive individuals remains at a constant level until the age of 30 , and from this age onwards a significantly higher level is observed ( p<0 . 001 ) . A relationship was observed between copro-Ag positivity and sero-Ag positivity ( p = 0 . 03 ) indicating that a copro-Ag positive individual was at an almost three-fold higher risk of being sero-Ag positive than the one who was not ( OR = 2 . 9 , p = 0 . 029 ) . There was no statistically significant difference in prevalence between males and females ( χ2 = 0 . 034 , p = 0 . 854 ) . Two clusters of cysticercosis cases ( 14 . 14048S , 31 . 19692E , density 31 . 31; 14 . 08460S , 31 . 22085E , density 195 . 17; 95th percentile density = 23 . 16 ) were identified in the study community with the larger cluster spatially related to the taeniosis cluster ( Figure 4 ) . Other intestinal parasites detected on coproscopic examination included hookworms , Schistosoma mansoni , Trichuris trichiuria and Ascaris lumbricoides . Table 2 shows the prevalence rates for these parasites with their respective 95% confidence intervals .
T . solium taeniosis tends to have a low prevalence , typically less than 1% , even in endemic communities [32] , a higher prevalence is considered hyper-endemic [33] . In this study a prevalence of 6 . 3% , based on copro-Ag ELISA , was determined , indicating a hyper-endemicity in the study community . As in a number of other studies , no significant association between age/sex and taeniosis positivity could be determined [31]–[35] . Even though similar high taeniosis prevalences have been recorded in Kenya ( 4–10% ) [17] , the 6 . 3% prevalence determined in this study should be looked at critically . The sensitivity and specificity of the copro-Ag ELISA are estimated at 96–98% and 98–100%; respectively [5] , [36] . However , the possibility of false positive test results due to cross-reactions with other pathogens present in the community should be considered . The assay has been reported not to cross-react with other parasite species including A . lumbricoides , T . trichiuria , Hymenolepis nana , H . diminuta or parasitic protozoa [36] . Also in our laboratory , stool samples with known H . nana , Schistosoma spp . , T . trichiuria and A . lumbricoides infections were analyzed , and all results remained under the cut off level ( Unpublished data ) . As the assay is not species specific [24] , the possibility of the high taeniosis prevalence to be partially due to T . saginata infections cannot be ruled out . However , bovine cysticercosis in Zambia has so far only been reported in the Central and Southern provinces [37] and Western province ( I . K . Phiri , personal communication ) . Interviews with local people in the study area revealed that pig slaughter and pork consumption increases in the dry season as it is time for harvest and residents have then the means to buy either an entire pig or pieces of pork for home consumption . During this period , pig owners not only slaughter more pigs for the market but also for their own home consumption . Higher pork consumption could have led to new ( still immature ) taeniosis infections at the time of sampling , which will be detected by copro-Ag ELISA but not yet by coprology [5] . Only one tapeworm ( from a participant positive on both copro-Ag and coproscopic examination ) could be recovered after treatment of the 45 copro-Ag positive participants . The low recovery rate of tapeworms can be explained by: ( 1 ) stools were obtained only over one day and not over 3 days post treatment [38] due to logistical constraints , ( 2 ) usually after antiparasitic treatment , small and unrecognisable fragments are expelled by most patients [38] and these are easily missed , and ( 3 ) treatment of copro-Ag positive individuals was conducted over six months after collection of stool samples; natural expulsion may have occurred in this period . The sero-Ag ELISA assay detected an apparent cysticercosis prevalence of 5 . 8% indicating the presence of viable cysts and as such active infections in these individuals . The prevalence of cysticercosis recorded in our study is comparable with that recorded in other endemic areas such as in the Andean region of Ecuador and in north Vietnam [39] , [40] , higher than in west Cameroon ( 0 . 4 to 4 . 0% ) and southern Ecuador ( 2 . 3% ) [41] , [42] but lower than that reported in the Democratic Republic of Congo ( 21 . 6% ) [14] . Other studies that have recorded higher seroprevalences include those that used antibody detection techniques such as in Mozambique ( 12 . 1% ) , South Africa ( 7 . 4% ) and Peru ( 13 . 9% ) [15] , [16] , [31] . However , antibody detection indicates exposure to the parasite and not necessarily established infection and hence is likely to detect more positives than the antigen detecting assay used in the current study [43] . Change point analysis of the association of antigen seropositivity and age revealed that the number of individuals in which circulating antigens were detected was significantly higher in people older than 30 years , indicating that viable cysts were more frequently present in individuals above this age . Studies have shown that a higher proportion of vesicular stage cysticercii is found in older ( 60 years and above ) NCC patients [44] , [45] and this has been attributed to immunosenescence since a weaker immunity in the elderly would facilitate the establishment and maintenance of viable cysticercii unlike in fully immunocompetent younger individuals [46] . The significant increase in sero-antigen positive individuals in the elderly was also observed in Ecuador where the number of positive individuals was higher in people order than 60 years [28] . However , in our study we see an increase already in young people ( from 30 years onwards ) who are supposed to be immunocompetent . Establishment and development of infection is influenced by a range of complex factors; among which are parasite factors ( e . g . parasite virulence/pathogenicity influenced by genetic differences , number , stage , location ) , host factors ( e . g . age , gender , genetics influencing the immunological responses of the host when exposed to infection ) and environmental factors ( e . g . presence of risk factors , level of exposure , presence of other infections ) [47] . It is very difficult to pinpoint exactly those factors present in the study area/population/age groups; that can explain this early increase in establishment of viable infection . The high taeniosis prevalence recorded in the study community entails a possible very high exposure risk to infective eggs . In a study in India , higher infection rates ( as indicated by sero Ag detection ) were noted in areas with higher taeniosis prevalences [48] . Also in our study a significant positive relationship between copro-Ag positivity ( presence of tapeworm ) and sero-Ag positivity ( cysticercosis ) was established ( logistic regression and cluster analysis ) . Level of exposure/infection with which the host is confronted can have an important effect on the immunological response of the host [44] , and as such on the establishment of viable infection . General factors that lower immunity in groups of individuals in the population could be at play making them more susceptible to infection . According to the United Nations Human Development Indices of 2008 , about 64% of Zambia's population lived on less than $1 per day as compared to only 20% for Ecuador [49] . Poverty is an indication for poor nutritional status , which has an impact on the immune system [50] . Also the presence of other diseases such as HIV-AIDS , malaria and tuberculosis , other helminthic infections and physical environmental conditions [51] can influence greatly the host's reaction to other infections . In 2008 , Zambia's HIV prevalence stood at 14 . 3% with the age group between 20 and 40 years being the most affected [52] . The country is also endemic for malaria [53] and helminthic infections are widely reported in rural areas , as reported in this study . Genetic polymorphism of the parasite is another important determining factor for the establishment and development of infection . Nakao et al . [54] have described a cluster of isolates from Asia , and another cluster from isolates from Latin America and Africa . However , genetic differences within a cluster ( within a continent/country/region ) need to be evaluated as well . Several Zambian isolates are currently being examined , and preliminary results indicate a high genetic variability ( Unpublished results , K . Kanobana ) , which might explain differences in development of infection between regions . We have , in this study , shown that T . solium taeniosis and cysticercosis are present in the study community . Many issues remain unclear and obviously more work is required to understand the many factors that contribute to the transmission dynamics of the parasite and disease development in endemic rural areas . Also the economic impact and burden of disease in rural pig keeping communities of Zambia needs to be determined .
|
Taenia solium taeniosis/cysticercosis is a zoonotic infection endemic in many developing countries , with humans as the definitive host ( taeniosis ) and pigs and humans as the intermediate hosts ( cysticercosis ) . When humans act as the intermediate host , the result can be neurocysticercosis , which is associated with acquired epilepsy , considerable morbidity and even mortality . In Africa , most studies have been carried out in pigs with little or no data in humans available . In this human study , conducted in a rural community in Eastern Zambia , prevalences for taeniosis and cysticercosis were determined at 6 . 3% and 5 . 8% respectively , indicating the hyperendemicity of the area . Cysticercosis infection was strongly related with age , with a significant increase in prevalence occurring in individuals from the age of 30 onward . A collected tapeworm was confirmed to be T . solium . Risk factors associated with the transmission and maintenance of the parasite such as free roaming pigs , households without latrines , backyard slaughter of pigs without inspection and consumption of undercooked pork were also present . The findings of this work have identified the need for further research in the transmission dynamics and the burden that this infection has on the resources of poor local people .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"infectious",
"diseases",
"public",
"health",
"and",
"epidemiology",
"epidemiology",
"neglected",
"tropical",
"diseases",
"parasitic",
"diseases"
] |
2012
|
Taenia solium Infections in a Rural Area of Eastern Zambia-A Community Based Study
|
The discovery of expression quantitative trait loci ( “eQTLs” ) can help to unravel genetic contributions to complex traits . We identified genetic determinants of human liver gene expression variation using two independent collections of primary tissue profiled with Agilent ( n = 206 ) and Illumina ( n = 60 ) expression arrays and Illumina SNP genotyping ( 550K ) , and we also incorporated data from a published study ( n = 266 ) . We found that ∼30% of SNP-expression correlations in one study failed to replicate in either of the others , even at thresholds yielding high reproducibility in simulations , and we quantified numerous factors affecting reproducibility . Our data suggest that drug exposure , clinical descriptors , and unknown factors associated with tissue ascertainment and analysis have substantial effects on gene expression and that controlling for hidden confounding variables significantly increases replication rate . Furthermore , we found that reproducible eQTL SNPs were heavily enriched near gene starts and ends , and subsequently resequenced the promoters and 3′UTRs for 14 genes and tested the identified haplotypes using luciferase assays . For three genes , significant haplotype-specific in vitro functional differences correlated directly with expression levels , suggesting that many bona fide eQTLs result from functional variants that can be mechanistically isolated in a high-throughput fashion . Finally , given our study design , we were able to discover and validate hundreds of liver eQTLs . Many of these relate directly to complex traits for which liver-specific analyses are likely to be relevant , and we identified dozens of potential connections with disease-associated loci . These included previously characterized eQTL contributors to diabetes , drug response , and lipid levels , and they suggest novel candidates such as a role for NOD2 expression in leprosy risk and C2orf43 in prostate cancer . In general , the work presented here will be valuable for future efforts to precisely identify and functionally characterize genetic contributions to a variety of complex traits .
Genome-wide association studies have uncovered numerous robust associations between common variants and complex traits , but only a minority of these can be traced to protein-altering polymorphisms [1] . It is likely that most of these associations result from non-coding variants . One hypothesis is that such variants modify cis-regulatory sequences and thereby change the expression levels of one or more target genes . Variance in gene expression plays essential roles in numerous important processes and is highly heritable in human populations [2] . Considering this , the discovery of genetic variants that have a functional impact on gene expression is a potentially powerful means to facilitate more accurate and robust identification of associations between variants and disease . Such discoveries may also provide mechanistic insight into otherwise anonymous genotype-phenotype correlations that often span many correlated variants across multiple genes . In large part due to this potential there has been recent substantial interest in the identification of expression quantitative trait loci ( eQTLs ) [3]–[10] . Regulation of gene expression in the liver is of particular interest given its vital roles in maintaining homeostasis and health , including synthesis of most essential serum proteins , the production of bile and its carriers , and the regulation of nutrients . The liver is also the predominant organ in xenobiotic metabolism , and it has been estimated that 75% of the 200 most widely prescribed drugs are eliminated from the body through liver metabolism [11] . Altered metabolism by genetic factors affects the systemic availability and residence time of xenobiotics and hence their toxic and pharmacologic effects [12] . While eQTL studies have made valuable contributions to genetic research ( e . g . , [13] ) , there exist several practical limitations to consider . First , most eQTL studies are conducted in immortalized , lymphoblastoid cell lines ( LCLs ) , which clearly have utility for the interpretation of human disease associations , particularly with immunity-related phenotypes [14] , [15] . However , the use of such cell lines potentially introduces artifacts associated with immortalization , subsequent passage , and growth conditions prior to harvest [16] . Second , eQTLs may exhibit spatiotemporal specificity [17] , [18] , presumably driven by polymorphisms located within tissue specific regulatory elements , and eQTL studies may be maximally informative for any given trait when conducted in a relevant , non-transformed cell type . Third , environmental factors and other , mostly hidden , confounding variables are known to significantly affect gene expression levels and measurements [19]–[23] . Fourth , most eQTL studies fail to provide replication on an independent set of samples with independent experimental assessment ( see [24]–[27] , [23] for exceptions ) . We sought to address these limitations , and conducted two independent eQTL studies and compared these results to a third , published study . Genetic analyses were performed using Bayesian regression [28] , [29] after controlling for age , sex , ancestry , and unmeasured confounding variables [20] . Using the UC liver panel as a ‘discovery’ cohort and the UW and Merck data as replication panels , we found that ∼30% of eQTLs identified at stringent thresholds failed to replicate in either of the two replication studies . We show that this is likely due to several factors , including SNPs in probes , but the effects of unmeasured confounding variables were particularly pronounced . We also found that reproducible eQTL associations were enriched near proximal promoters and 3′ UTRs . Through targeted resequencing and luciferase experiments , we identified 3 significant haplotype-specific in vitro functional effects that directly support a liver eQTL . These data functionally validate the enrichment for eQTLs near gene ends and suggest that many eQTLs can be rapidly fine mapped to a causative variant or haplotype . Finally , given our study design we identified hundreds of genes with reproducible SNP-associated expression levels , a subset of which provide strong mechanistic hypotheses for published associations between SNPs and disease .
We analyzed two independent sets of primary liver tissues at the University of Chicago ( UC; n = 206 ) and University of Washington ( UW; n = 60 ) . We genotyped both sets of samples using Illumina SNP arrays ( quad-610 and 550 k for UC and UW , respectively ) ; to improve mapping power [30] , [28] and replication ascertainment , additional genotypes were imputed using HAPMAP reference genotype panels ( see Methods ) . Gene expression levels were analyzed using Agilent ( UC ) and Illumina ( UW ) expression arrays . We considered the UC liver collection as a ‘discovery’ set and used as replication panels the UW collection and a published set of liver eQTL data ( Merck; n = 266 ) [31] . However , we note that the conclusions drawn below were robust to the choice of a ‘discovery’ set ( Figure S1 ) . All samples analyzed across all three studies were unique . Microarray expression probes from both platforms were remapped to RefSeq gene models to aid in cross platform comparisons . A total of 14 , 703 RefSeq genes were surveyed in the UC reference study while 11 , 245 RefSeq genes were present on all three platforms . We have made these data and results publicly available through the GEO and SCAN databases ( http://www . scandb . org/ ) [32] . After correcting for technical effects and unmeasured confounding variables , we found that thousands of gene expression traits were significantly associated with demographic variables . At a 5% false discovery rate ( FDR ) , 769 , 336 , and 3 , 110 genes were significantly associated with ancestry , sex , and age , respectively within the UC livers . Genes significantly affected by sex or age ( FDR<5% , Figure 1 , Figure S2 , examples displayed in Figure S3 ) have a marked enrichment for small p-values in both replication samples ( Figure 1A , 1D ) . To lessen the influence of differential statistical power among the three studies ( n = 206 , 60 , 266 ) , we defined ‘replication’ as having a nominally significant p-value in the independent sample ( p-value<0 . 05 ) and having a concordant effect direction ( i . e . , is YFG more highly expressed in males or females ? ) . 29 . 9% and 32 . 1% of genes significantly affected by sex ( UC sex t-test FDR<5% ) replicated in the UW and Merck studies , respectively ( Figure 1B ) . At more stringent thresholds , validation rates exceeded 80% , albeit with fewer included genes ( Figure 1B ) . We also note that the sex-associated gene set was strongly enriched for genes on the X and Y chromosomes ( Figure 1C; X chromosome , hypergeometric test , p-value = 1 . 72×10−14 , Y chromosome , p-value<2×10−16 ) , as would be expected for genes with sex-associated expression levels . Effects due to age were less reproducible: 13 . 2% and 21 . 5% of significantly age associated genes ( UC age t-test FDR<5% ) replicated in the UW and Merck studies , respectively ( Figure 1E; an example of a replicated age-associated gene , TMEM22 , is displayed in Figure 1F ) . Effect sizes for both sex and age were correlated across studies ( Figure S4; Spearman's rho , UC-UW sex = 0 . 597 , UC-Merck sex = 0 . 720 , UC-UW age = 0 . 333 , UC-Merck age = 0 . 159 ) , underscoring the reproducibility of demographic effect estimates . It is possible that both age and sex replication rates are downwardly biased due to differences in age and sex distributions ( Table 1 ) . To quantify the potential effects of heterogeneous sample sizes and unbalanced study designs , we performed resampling studies within the UC discovery cohort . Demographic effect replication rates were recalculated using 60 samples that were race , sex , and age ( +/−3years ) matched to the UW distribution ( Figure 1B , 1E; see Methods ) . We found that 34% of sex effects and 15% of age effects replicated by simulation , supporting the conclusion that sample size and demographic heterogeneity do generate significant covariate associations that our replication studies are unable or underpowered to detect . After adjusting for age , sex , ancestry , and unmeasured confounding variables ( quantified by surrogate variable analyses , see Methods and [20] ) , we found 1 , 787 gene models with significant cis-linked genetic effects on expression levels ( UC log10 Bayes Factor ( BF ) >5; SNP to TSS distance <250 kb; Figure 2A , Figure 3A , Table S1 ) . The distribution of t-test p-values in the replication sets , adjusted for the same covariates , for the UC best associated gene-SNP pairs were significantly enriched for small values ( Figure 3B ) , indicating that a large fraction of cis-eQTLs are reproducible in independent sample collections . As with demographic effects , we defined replication as a p-value<0 . 05 and a concordant allele effect direction ( Figure 3C ) . While the significance of association in the discovery cohort has a large effect on replication probability , the relationship between significance and replication was effectively binary ( Figure 3C ) . Cis-eQTLs with BFs>5 were much more likely to replicate than those with BFs<5 ( chi square p-value<2×10−16 ) . However , among genes with BFs>5 , replication probability was only weakly dependent upon BF ( Figure 3C; logistic regression chi-squared p-value = 0 . 00319 ) . We found that 49 . 1% and 57 . 6% of significant cis-eQTLs ( UC BF>5 ) replicated in the UW and Merck studies , respectively ( i . e . , p-value<0 . 05 and concordant effect directions; Figure 2A–2B , Figure 3C ) . The lower observed replication rate for the UW study is partially attributable to the smaller sample size ( 60 vs 266 ) , but may also reflect platform-dependencies . 66 . 7% of significant cis-eQTL associations replicated in at least one of the two replication cohorts , while 47 . 6% replicated in both cohorts . Cis-eQTLs that replicated in one replication study were significantly more likely to replicate in the second replication study than expected by chance ( chi-squared p-value<2×10−16 ) and twice-replicated eQTLs had larger effect sizes than eQTLs that replicate in only one study ( Wilcoxon rank-sum test p-value<2×10−16; Figure S5 , examples of non-replicated cis-eQTLs displayed in Figure S7 ) . Given differences in sample sizes among these studies , we sought to define a baseline replication rate against which to compare the observed levels of reproducibility . We therefore conducted a re-sampling experiment in which , for each gene expression trait , 100 sets of 60 sex and age ( +/−3 years ) matched samples were selected at random and used to define replication ( i . e . concordant effect direction and p<0 . 05 ) . We found that simulated replication rates increase dramatically near a BF of 5 ( 95 . 5% replication at BF>5; Figure 3C ) and are effectively 100% at higher thresholds . These observations suggest that power differential among the studies cannot alone explain the observed rates of replication , as there are many genes with effect sizes in one study that should be readily detected in both ( let alone either ) replication panels . This is further supported by the observation that concordance alone ( i . e . no p-value threshold ) yielded similar levels of reproducibility , as did direct comparisons of allelic coefficients ( Spearman's rho of 0 . 663 and 0 . 681 for UC–UW and UC-Merck comparisons , respectively; Figure S6 ) . We next sought to evaluate whether ‘winner's curse’ [33] , [34] was deflating replication rates . Therefore , we extracted simulations in which the estimated coefficients randomly decreased and found that simulated replication remained >90% at BF>5 and near 100% at higher BF even when the effect size declined substantially ( e . g . 30% drop in regression coefficient; Figure S9 ) . Effect sizes would need to be over-estimated by 2-fold or greater across the entire set of eQTLs with UC BF>5 to result in the observed rates of replication . Furthermore , two lines of reasoning suggest winner's curse is not a major contributor to the observed rates of non-replication . First , we note that bias resulting from winner's curse should be progressively less pronounced as the true effect size increases , which in turn will correlate with significance estimates in the discovery panel [34] . However , replication rate was essentially flat even at extremely stringent thresholds ( Figure 3C ) . Additionally , the resampling experiments demonstrated that , in direct contrast with a winner's curse prediction , effect sizes would need to be increasingly more severely over-estimated at higher thresholds ( 3-fold or more ) to result in the observed rates of replication ( Figure S8 ) . Second , the definition of replication ( concordance and p-value<0 . 05 ) is relatively loose when applied to eQTLs with a BF>5 ( typical linear regression p-values<5×10−8 ) and should accommodate substantial drops in effect sizes for both replication panels but especially for the larger Merck dataset . This is further supported by the observation that concordance alone yielded similar rates of replication ( Figure S6 ) . We conclude that statistical power and winner's curse cannot explain the observed rates of non-replication for eQTLs with BF>5 . One possible explanation for non-replication is that SNPs within sequences targeted by expression probes may change hybridization efficiency in an allele-specific manner; if that SNP is also correlated with a genotyped variant , false positive eQTLs may result [35] . While 45 . 3% and 37 . 2% of Agilent and Illumina probes overlap with a polymorphism found in dbSNP131 or the one thousand genomes project ( 2010 . 08 . 04 release ) , the frequency distribution of polymorphisms in and around probe sequences differs markedly between the Agilent ( UC ) and Illumina ( UW ) platforms ( Figure S9 ) ; Illumina expression probes have clearly been designed to avoid common polymorphisms . The presence of SNPs in expression probes had a larger effect on reproducibility at extremely high thresholds ( Figure 3C ) . For example , the replication rate for cis-eQTLs with BF>5 is not significantly affected by the presence of SNPs in probes ( p-value = 0 . 189 ) ; however , replication rate for cis-eQTLs thresholded at BF>10 is significantly affected by probe SNPs ( p-value = 0 . 0354; 65 . 6% with , 74 . 9% without SNP ) and replication rate is significantly associated with an interaction between probe SNPs and eQTL significance ( logistic regression BF-SNP interaction p-value = 0 . 0224 ) . These results suggest the proportion of non-reproducible cis-eQTLs increases with eQTL significance such that , for eQTLs with BF>10 , ∼27% of the non-replication rate can be explained by the presence of hybridization artifacts caused by known polymorphisms . To investigate the potential confounding role of unannotated polymorphisms in eQTL ascertainment , we re-sequenced 15 expression probes for genes that had large discrepancies in correlation measurements between the UW and UC studies that did not overlap a known SNP ( 9 probes with strong UW correlation but low UC correlation , 6 of the converse; Table S2 ) . We found that none of these 15 probes harbored SNPs in the 60 UW liver samples or a panel of 35 CEU HapMap samples . Collectively , our data suggest future array designs/eQTL studies would benefit from more aggressive avoidance of known SNPs , but current SNP annotations are sufficiently comprehensive that unknown variants are of little concern to eQTL analyses . We next quantified the role of several additional factors that may generate spurious associations . Most strikingly , failure to control for unknown or unmeasured confounding variables by surrogate variable analysis ( SVA ) produced a large decrease in the number of significant ( BF>5 ) cis-eQTL signals ( 1 , 787 vs . 873; Figure 4A; McNemar's chi-squared test p-value<2×10−16 ) , similar to a recent study of gene expression within twins [23] . Not only did SVA produce a larger number of significant cis-eQTL associations , but these associations were also significantly more likely to replicate ( McNemar's Chi-squared test p-value≪2×10−16; Figure 4B ) . While it has been shown that unknown or unquantified confounders can lead to unreliable genetic predictions [19] , [36] , [2] , our data show that such factors , if unaccounted for , dramatically decrease the number of eQTL signals and their reproducibility across multiple independent collections of primary human tissues . Several additional aspects of the gene expression measurements correlated with cis-eQTL replication rate . Cis-eQTL replication rate was significantly associated with mean gene expression level and , independently , inter-individual expression coefficient of variation ( Figure S10; multivariate logistic regression chi-squared p-value = 3 . 44×10−3 and 1 . 41×10−4 , respectively ) ; more highly expressed and highly variable genes were more likely to replicate . Further , we found that expression variance unexplained by age , sex , race , and surrogate variables was negatively correlated with expression level ( Spearman's rho = −0 . 302 , p-value<2×10−16 ) . These data suggest greater measurement accuracy at higher expression levels that leads to more robust eQTL identification . We also found that the best associated SNP for each gene expression trait was frequently immediately upstream or downstream from the transcription start site ( TSS ) ( Figure 2C , [37] ) . Replication rate of significant cis-eQTLs was associated with absolute SNP to TSS distance ( logistic regression chi-squared p-value = 5 . 35×10−3 ) . 74 . 5% of cis-eQTLs within 5 kb of the TSS replicated , compared with only 60 . 6% located more than 100 kb from the TSS . Thus while distal regulatory elements are clearly important for human gene expression regulation , robustly quantifiable , segregating expression polymorphism was more likely to be found in SNPs very close to the TSS of genes . Interestingly , significant cis-eQTLs were no more likely to replicate when analyses were restricted to probes targeting the same exon ( chi-squared p-value = 0 . 759 ) , demonstrating that most non-replicating eQTLs ( in our study design ) can not be accounted for by differential splicing or isoform usage . Similarly , replication was not improved when analyses were restricted to gene expression measurements derived from more than one expression probe ( chi-squared p-value = 0 . 919 ) . Additionally , the minor allele frequency of the associated SNP did not have a significant effect on replication rate ( logistic regression chi-squared p-value = 0 . 600; Figure S10 ) , and eQTLs at imputed SNPs replicated at similar rates to directly genotyped SNPs ( logistic regression chi-squared p-value = 0 . 574; Figure S10 ) . Uncertainty at imputed SNPs does not appear to have a significant effect on cis-eQTL replication rate , as the ratio of observed to expected genotype variance was not associated with replication rate in any of the three sample sets ( logistic regression chi-squared p-values all >0 . 152; Figure S12 ) . Examination of the interplay of the factors influencing eQTL replication revealed several interesting trends . As mentioned above , replication probability was significantly associated with SNP to TSS distance , but this association decreases with increasing cis-eQTL significance ( distance×BF interaction logistic regression p-value = 3 . 98×10−5 ) . Thus , location information can help to differentiate real from false positive correlations of modest effect , but is less important for very strong correlations . We constructed stepwise multivariate logistic regression models , restricted to associations with BF>5 , and confirmed that BF ( logistic regression chi-squared p-value = 7 . 32×10−3 ) , SNP to TSS distance ( p-value = 2 . 33×10−3 ) , gene expression ( p-value = 0 . 0230 ) , gene expression CV ( p-value = 1 . 33×10−4 ) , and probe SNP×BF interaction ( p-value = 0 . 0207 ) all have significant effects on the cis-eQTL replication rate . In contrast , SNP minor allele frequency , SNP type ( imputed or direct ) , and genotype variance do not substantially influence replication rate ( p-values>0 . 5 ) . We also conducted genome-wide scans for associations between gene expression traits and unlinked SNPs . Such trans-eQTLs may represent regulatory interactions between transcription factors , signaling molecules , or chromatin regulators and their target genes . After adjusting for demographic variables as above , we found 353 gene expression traits with significant ( BF>5 ) trans-linked genetic effects . The replication behavior of trans-eQTLs was markedly different from cis-eQTLs ( compare Figure 3B , 3C with Figure 3E , 3F ) . First , the distribution of t-test p-values derived from the UW replication set , for each best associated gene-SNP pair identified in the UC set , was effectively uniform ( Figure 3E ) . Second , in contrast to cis-effects , which rapidly approach an asymptotic replication rate at BF 5 , trans-eQTLs almost completely failed to replicate ( 6 . 14%; Figure 3F ) at a BF threshold of 5 . At greater significance thresholds , trans effects did replicate more frequently ( e . g . , at BF> = 9 . 5 , 50 . 0% replicate ) , however , these rates never approached those observed for cis-eQTLs . It is plausible that surrogate variable correction may mask true ‘master’ regulator effects , but as for cis-effects we identified more trans-eQTLs with surrogate variable correction than without and these associations were more likely to replicate ( data not shown ) . While it is perhaps surprising that even extremely significant trans effects frequently fail to replicate , we note that this behavior is , to a certain extent , to be expected [27] . As the eQTLs we identified are associations between effectively anonymous SNPs and expression of a nearby gene , we were also interested in fine-mapping the associations , ideally to a causal variant ( expression quantitative-trait-nucleotide or eQTN ) or haplotype . We therefore re-sequenced the promoter and 3′UTR sequences for 18 genes with strong cis-eQTLs within the 60 UW livers ( Table S3 ) . Thirteen of these genes harbored a common SNP or indel within the proximal promoter or 3′UTR that correlated strongly ( p-value<1×10−8 ) with the expression level of that gene , while 17 of 18 harbored a variant with at least a modest correlation ( p-value<0 . 001 ) . Of these 17 genes , the most strongly correlated SNP was within the 3′UTR for 11 genes and within the promoter region for 6 genes . Moreover , 10 of the 17 best SNPs were not within HapMap , indicating that a majority of the most strongly associated promoter/3′UTR variants were neither genotyped directly nor imputed and therefore not detectable in the original eQTL analysis . We subsequently sought experimental support for the functional nature of the most strongly associated SNPs . Therefore , for 14 genes , we cloned ( and sequence-verified ) common haplotypes existing in the UW liver samples into a customized luciferase reporter vector , and tested the function of each haplotype using high-throughput , transient transfection reporter assays ( Table S3; 9 of 14 underlying cis-eQTLs replicated in the UC or Merck samples ) . For each haplotype , multiple independent vector ( mode of 3 ) preparations were made , and for each plasmid preparation 4 transfection replicates were performed ( mode of 12 measurements per haplotype ) . We analyzed the resulting data using a random-effects model that accounted for both variation in transfection replicates and variation in vector preparations . Our results underscore the need to perform multiple independent DNA preparations to reliably infer sequence-specific functional effects with this system ( Figure 5 and data not shown ) . We identified three regions where the haplotype sequence had a significant ( p-value<0 . 001 ) effect on reporter activity ( luminescence ) in the same allelic direction as the expression measurements , including two promoters and one 3′UTR region ( Figure 5 and Figure S12 ) . No significant but discordant effects were observed . Variants near PRMT6 , which encodes a protein-arginine methyltransferase and has been associated with HIV infection progression [38] , scored highly in both the UW and UC eQTL analysis ( Figure 5A ) . Resequencing of the PRMT6 promoter yielded two common haplotypes defined by two perfectly correlated SNPs located 406 and 150 bp upstream from the TSS . The minor haplotype ( 40% frequency ) correlated with a strong additive decrease in PRMT6 liver expression ( t-test p-value = 6 . 4×10−14 for UW ) , and relative to the major haplotype , we found a concomitant decrease in luminescence for reporter constructs harboring the minor haplotype ( p-value = 0 . 0002 ) . A similar result was obtained for promoter haplotypes of the LDHC ( lactate dehydrogenase C ) gene in which six common variants defined 7 common haplotypes , five of which were successfully cloned and tested . The strongest expression correlation was observed for a SNP located 392 bp upstream of the TSS ( 15% MAF ) , and the luciferase data strongly support the functional effect of this variant ( p-value = 8 . 7×10−9; Figure 5B ) . Finally , we identified a significant haplotype-specific effect within the 3′UTR for IPO8 ( importin 8 ) , a protein that interacts with Argonaute proteins to direct miRNA mediated gene expression regulation [39] . There were nine common 3′UTR haplotypes defined by 13 variants for IPO8 . The two haplotype groups defined by the most strongly expression-associated SNPs ( two perfectly correlated variants at positions 1147 and 1195 relative to the 3′UTR start ) have significantly different ( p-value = 9 . 5×10−4 ) functional effects . However , unlike LDHC , there remained a substantial amount of variance within the haplotypes defined by alleles at these two SNPs , suggesting other variants may also have a functional role . Alternatively , the data gathered from 3′UTRs were generally noisier than that for promoters ( Figure 5 and data not shown ) , and may not be as sensitive for identifying sequence-specific 3′UTR effects . Due to the increased noise , we repeated the analysis and performed new clone preparations and transfections for a subset of the IPO8 haplotypes . The replicate data also show a significant ( p-value = 0 . 007 ) difference , in the same direction , between haplotypes defined by their 1147 ( or 1195 ) allele ( Figure S12 ) .
Genetic analyses of gene expression have great potential to facilitate insights into the genetic basis of complex traits . However , the utility of these data are limited by the extent to which the discovered associations correspond to legitimate , reproducible associations . Our estimates of 49% ( UC vs . UW ) , 57% ( UC vs . Merck ) , and 67% ( UC vs . either ) cis-eQTL reproducibility are substantially lower than two recent reports between two mouse crosses ( 76% , [27] ) , two independent sets of lymphoblastoid cell lines ( 83% , [25] ) , and two sets of primary human skin ( >99% , [26] ) . Several non-exclusive possibilities likely contribute to these discrepancies . First , different discovery methodologies and replication criteria were employed in each study . Second , our studies were performed on different expression platforms ( Agilent and Illumina ) , which reduces the influence of reproducible platform-specific errors but may result in missing splice-variant-specific eQTLs [40] , [41] , [10] as array manufacturers often target different exons in a given gene . However , this is likely to have a limited effect , as we found that the replication rate was not significantly different for genes assessed by probes within the same exon ( Figure S10 ) . Third , we compared three independent collections of primary human tissues ( see Methods ) , not transformed cell lines or mouse tissues , and , despite the interpretive advantages associated with the former , our replication rate estimate is possibly downwardly biased by cell type heterogeneity . Finally , other systematic differences between studies , including protocols for sample collection and storage , clinical interventions taken by patients prior to death and autopsy , causes of death , life histories , etc . , may contribute to non-reproducibility . This hypothesis is supported by the observation that drug exposures and other clinical covariates , for which data limitations prevent comprehensive analysis , have substantial effects on gene expression; for example , we found that drug metabolism genes were significantly up-regulated in barbiturate-exposed vs non-exposed livers ( data not shown ) . The striking difference in reproducibility between the results reported here and a recent report quantifying the overlap of human skin eQTLs [26] , suggests that the degree of functional tissue heterogeneity may vary substantially across tissues . An important caveat is that these estimates of reproducibility are less meaningful for sequence-based studies of gene expression , which offer advantages in dynamic range and measurement accuracy [9] , [10]; sequencing is also largely immune to the SNP-in-probe effect that significantly inflates false positives in our data ( Figure 3C ) . However , the observation that age , sex , race , drug exposures , clinical covariates , and other global factors have such strong influences on expression ( e . g . , this study and [36] , [2] ) coupled with observations in other studies and different tissues that factors like cause of death are relevant [42] , suggests that much of the non-reproducibility is in fact driven by systematic differences in tissue source . Such differences will likely be important to all studies of primary tissue samples , whether assayed by arrays or by sequencing . The reproducibility of future results would benefit from analysis of samples from multiple centers with as much clinical information as possible . Furthermore , our results confirm previous observations that the effects of unknown , unmeasured , or unquantified covariates can confound genetic effects with structured error sources [19] , [36] , [2] , and that controlling for these hidden confounders substantially boosts the rate of eQTL discovery [23] . Importantly , we demonstrated that not only are more eQTLs detected but that their reproducibility in independent collections of primary human tissue was also significantly higher . Finally , through resequencing and a widely used in vitro assay system [43] , we found that of 14 tested genes , two genes harbored functional eQTNs in the proximal promoter and one gene harbored functional eQTNs in the 3′UTR . The success rate of 3 in 14 suggests that a substantial number of eQTLs , and by extension any complex traits that they may influence , can be functionally isolated using the scalable assay system that we employed or potentially higher-throughput assays [44] . We note that some truly functional variants will not be detectable in these assays , either from being tested out of their genomic context or having effect sizes below the limit of detection afforded by the number of replicates used ( e . g . [45] ) , and that the actual fraction of eQTLs with promoter or 3′UTR functional variation may be substantially higher . Considering that replication was significantly greater for eQTLs near the ends of genes relative to those further away ( Figure 3D ) , our functional analysis also strongly supports the use of SNP to gene distance as an important contributor to the prior probability that any given SNP is a cis-eQTN [37] . While some eQTNs clearly reside outside these regions ( e . g . , [46] ) , the heavy enrichment for reproducible and experimentally tractable eQTNs , coupled with historical evidence supporting disease relevance [47] , [48] , suggests that the relatively small ‘promoter’- or ‘3′UTR’-ome target spaces may be valuable additions to exome-based disease resequencing efforts [49] . Given the ubiquitous importance of gene expression variance to phenotype , the known heritability of gene expression variance , and the great preponderance of non-coding functional elements in the genome [50] , complex disease studies can benefit from eQTL analyses . Towards that end , we searched for correlations between replicated eQTL SNPs identified here and complex trait associated SNPs ( R2>80%; Table 2 , Table S4 ) in the NHGRI GWAS catalog ( http://www . genome . gov/gwastudies/ ) . These included several previously characterized mechanistic links to complex traits , such as VKORC1 expression and warfarin drug response [51] and SORT1 expression correlations with lipid levels and heart disease [13] , both of which were originally identified using the UW liver panel described here . Additionally , these data support a relationship , which had previously been speculated but not shown to exist , between NOD2 expression levels and leprosy risk [48] , and novel hypotheses such as a link between expression of the uncharacterized C2orf43 gene and prostate cancer risk [52] . In summation , our data facilitate insights into the factors and experimental design criteria that affect eQTL reproducibility and may improve future eQTL studies , replicate many published but nonreplicated eQTLs ( e . g . from [31] ) , support and extend eQTLs identified in other tissues like brain ( e . g . FAM119B [53] ) , identify many novel reproducible liver eQTLs , show that promoters and 3′UTRs are enriched for experimentally accessible functional variation , and support or suggest numerous mechanistic links to biomedically important phenotypes . We believe that this study and others like it will be valuable to the robust discovery and fine-mapping of the genetic basis for complex human diseases .
Research conducted in this study was performed on deceased , anonymous individuals and is therefore not considered to involve ‘human subjects . ’ Samples were collected with approval of institutional review boards ( IRBs ) and the University of Chicago and University of Washington IRBs approved their use for the purpose of this study . Livers were processed through Dr . Mary Relling's laboratory at St . Jude Children's Research Hospital , part of the Pharmacogenetics of Anticancer Agents Research ( PAAR ) Group , and were provided by the Liver Tissue Cell Distribution System funded by NIH Contract #N01-DK-7-0004/HHSN267200700004C and by the Cooperative Human Tissue Network . Samples were collected with approval of institutional review boards ( IRBs ) and the University of Chicago IRB has approved their use for the purpose of this study . Analysis began with 240 normal ( non-diseased ) livers that were collected from unrelated donors of self-reported European and African descent . Most of the liver tissue samples come from donor livers that were not used for whole organ transplants , the remainder being from liver tissue which remains following a partial graft into a smaller recipient , usually a pediatric patient . As such , each liver is procured with the intent to transplant under the best possible conditions to maintain cell viability . Standardized procedures have been in place for handling , freezing and storage of the livers and their subcellular fractionation and enzyme characterization . Demographic information is summarized in Table 1 . The University of Washington IRB approved the collection of the liver tissues and their subsequent use for the purposes of this study . Samples of human liver were obtained from organ donors through the University of Washington Transplant Program and the Northwest Organ Procurement Agency . Consent for research was obtained in all cases . Standard procedures were employed for the handling , freezing and storage of the livers . Gene expression microarray experiments were conducted with biological replication in all samples . Sample processing order was randomized . For each sample , total RNA was extracted at least twice independently , from tissue homogenized in TRIzol reagent , followed by Qiagen RNAeasy cleanup ( Qiagen ) . RNA quality was assessed by Bioanalyzer ( minimum RIN = 7 ) . cRNA was produced using the Agilent Low-Input Linear amplification and labeling kit . Array hybridizations ( Agilent-014850 4×44 k arrays , GPL4133 ) were performed at The University of Chicago , Argonne National Labs high throughput genome analysis core facility , according the manufacturers instructions . The Agilent FE software was used to extract feature intensities and to flag saturated , non-uniform , and outlier features . Probe intensity was adjusted by subtracting background intensity using the minimum method [54] , [55] and quantile normalized between arrays [56] . Dixon's outlier test was used to remove 13 arrays ( out of a total of 517 ) based on total number of flagged probes , intra-array variance , inter-array variance , biological replicate variance , and spike-in linearity [57] . Probes were grouped into probe sets by aligning first to RefSeq gene annotations and then aligning unmapped probes to the human reference genome ( build 36 ) . All probes with non-unique best alignments were excluded from further analysis . Multiprobe probesets were hierarchically clustered using one minus the pearson correlation coefficients as a distance matrix . Clusters were divided into groups by cutting clusters at a dendrogram height of 0 . 5 ( roughly producing clusters with internal correlation coefficients >0 . 5 ) . All downstream analyses were performed independently on each resulting cluster and all single probe probesets . Total RNA was extracted from 60 human liver tissue samples from the University of Washington School of Pharmacy Human Liver Bank as previously described [51] , [13] . Genome wide expression analysis was performed using 750 ng of total RNA on the Illumina HumanRef-8 v . 2 platform ( GPL5060 ) . All liver samples were analyzed with technical replicates that were randomized between processed batches of 24 arrays performed on different days . Raw signal intensity measurements from each sample were processed using the Illumina BeadStudio software v . 2 . 3 . 41 using the ‘average’ normalization function . Replicate data from each liver was averaged prior to statistical analysis . All samples and replicates passed quality-control measures . Processed gene expression data from the published Merck liver eQTL study [31] were downloaded from GEO ( GSE9588 , GPL4371 ) . Based on available sample metadata , 266 samples had ( a ) unambiguous sample ID , age and sex assignments ( b ) expression data , ( c ) genotype data , and ( d ) did not overlap with the UC study . Probes were grouped into RefSeq gene annotation probe sets based on the array manifest . Probesets were further clustered and split following the methodology used for the UC array set . From the same liver samples received from the Liver Tissue Resource , DNA was obtained from 240 samples for genotyping . Genotyping was performed on the Illumina human 610 quad beadchip platform ( GPL8887 ) at the Northwestern University Center for Genetic Medicine Genomics Core Facility according to the manufacturer's instructions . One sample was removed because it had a no call rate >10% . The initial marker set comprised 620 , 901 markers . 8 , 300 markers were removed because they showed significant deviation from Hardy-Weinberg equilibrium ( HWE , Fischer's exact test , p<0 . 001 ) . 29 , 705 SNPs were removed from the analysis because they had a no call rate in more than >10% of the samples . Hence , our final marker set is comprised of 583 , 073 SNPs . Identity by descent analysis , performed in Plink , revealed 14 pairs of duplicated samples . Erroneous , redundant sample collection was later confirmed by the tissue bank . Genotype and expression data for these samples were merged for all downstream analyses . The final sample set therefore consisted of 225 unique samples . Genotyping was performed on each liver sample using the Illumina HumanHap550 ( GPL6981 ) Beadchip platform . Genotyping calls were made using GenomeStudio . After raw genotyping data were loaded into the software , pre-defined cluster definitions were applied and genotype calls were determined . Clusters were checked for separation , deviation from HWE , and lack of variation ( i . e . , monomorphic ) . Poorly assigned clusters were modified manually and sites were re-called with corrected cluster definitions . All samples had call rates greater than 98% . Genotype data were generated as described [31] . The sex of each sample was imputed by K-means clustering of Y-linked gene expression levels and X- and Y-linked genotypes . 3 UC samples , 0 UW samples , and 0 Merck samples had mismatched imputed and annotated sexes , and were therefore excluded from all analyses . For all three studies , care was taken to translate all genotypes to reference genome ( b36 ) forward strand alleles , as subtle errors in genotype strand inference will downwardly bias replication rate estimates . Additional genotypes were imputed with Bimbam ( v 0 . 99 ) [58] , using HAPMAP release 27 , build 36 unphased genotypes as reference panels . European American genotypes were imputed with a CEPH reference panel , while African American genotypes were imputed with a combined CEPH and YRI panel . Imputation was run with default Bimbam parameters , and mean imputed genotypes were recorded and used for all downstream analyses . We performed a principal component analysis ( PCA ) based quantification of race using the African and European populations from the Human Genome Diversity panel as reference populations . The SNP set was trimmed using linkage disequilibrium ( LD ) -based SNP pruning , removing all SNPs for with high pairwise LD ( R2>0 . 8 ) , as in [59] . PCA was performed using smartpca , as implemented in EIGENSOFT [60] . Four samples were flagged as outliers and removed from all further analyses . As expected , the first principal component separated African from non-African individuals . We therefore used this loading vector as an estimated quantification of African ancestry for further analyses . PCA was performed using the multi-dimensional scaling procedure implemented in PLINK v1 . 06 ( http://pngu . mgh . harvard . edu/purcell/plink/ ) [61] . The vast majority of samples resided in a single cluster including all the individuals of self-reported European ancestry , with several moderately outlying samples corresponding to self-reported Hispanic and African ancestry . No samples were excluded from further analyses; the vectors determined for the first two principal components were used as ancestry control for all statistical analyses . All 266 samples included from the published Merck study were self-reported Caucasians . The SNP set was trimmed using linkage disequilibrium ( LD ) -based SNP pruning , removing all SNPs for with high pairwise LD ( R2>0 . 8 ) , as in [59] . PCA was performed using the multi-dimensoinal scaling procedure implemented in PLINK v1 . 07 ( http://pngu . mgh . harvard . edu/purcell/plink/ ) [61] . No outliers were detected; the vectors determined for the first four principal components were used as ancestry control for all statistical analyses . For each probeset , surrogate variable analysis ( SVA ) [20] was performed on the matrix of expression measurements , after controlling for the effects of hybridization protocol , age , sex , and a principal component analysis based quantification of genetic ancestry . For each probeset , we then constructed a linear mixed effects model y ∼ m + P + A + C + R + I + W + SVi . . n + e , where y is the log2 transformed probe intensity , m is the expected probe intensity , P is a factor controlling for the effect of subtle variations in hybridization protocol ( e . g . , the identity of the technician who performed the experiment ) , A is the effect of individual age , and C is the effect of individual sex , and R is the effect of genetic ancestry . I is the random effect of each individual , W is the random effect of the oligonucleotide probe , SVi . . n represents the effects of a matrix of 55 surrogate variables , and e is the residual error . The model was fitted to each gene by residual maximum likelihood using the lmer function in the R package lme4 ( v 0 . 999375-32 ) [62] , [63] . Fixed effect p-values were estimated using the pvals . fnc function in the languageR package ( v 1 . 0 ) [64] . The significance of covariate effects was assessed by estimating false discovery rates , using Storey's q-value method [65] . To further control for the effects of outliers and population stratification , prior to eQTL mapping , the distribution of estimated individual effects , for each gene expression trait , was normal quantile transformed , within populations . SVA [20] was performed on the matrix of expression measurements , after controlling for the effects of age , sex , and a multidimensional scaling based quantification of genetic ancestry . For each probe , we constructed a linear model y ∼ m + A + C + R + SVi . . n + e , where y is the log2 transformed probe intensity , m is the expected probe intensity , A is the effect of individual age , and C is the effect of individual sex , and R is the effect of genetic ancestry , SVi . . n represents the effects of a matrix of surrogate variables , and e is the residual error . Models were implemented with the lm function in R . The residuals from this regression were used as the phenotype values for all subsequent analyses . SVA [20] was performed on the matrix of expression measurements , after controlling for the effects of age , sex , and a principal component analysis based quantification of genetic ancestry; 54 significant surrogate variables were identified . For each probeset , we then constructed a linear model y ∼ m + A + C + R + W + SVi . . n + e , where y is the log2 transformed probe intensity , m is the expected probe intensity , A is the effect of individual age , and C is the effect of individual sex , and R is the effect of genetic ancestry , W is the effect of the oligonucleotide probe , SVi . . n represents the effects of a matrix of surrogate variables , and e is the residual expression . Models were implemented with the lm function in R . The residuals from this regression were used as the phenotype values for all subsequent analyses . For each gene expression trait , residual expression variance was treated as a quantitative trait and tested for association with all markers genome-wide . Association testing was performed by Bayesian regression , as implemented in Bimbam ( v 0 . 99 ) , using mean imputed genotypes and default priors [28] , [29] . Genotypes with minor allele frequencies less than 1% were excluded . For 15 probes that showed discrepant eQTL scores between the UC and UW analyses ( i . e . BF>4 in one study and BF<4 in the other ) , we designed primers to capture the relevant expression array probe and amplified and Sanger-sequenced the resulting PCR products in each of the 60 UW liver samples and 35 CEU HapMap samples . SNPs were identified as previously described ( http://pga . gs . washington . edu/ ) including both automated prediction and manual curation . We resequenced the promoter and 3′UTR regions within the 60 UW liver samples and 35 CEU HapMap samples for 18 genes that showed strong expression-SNP correlations within the UW data ( selected before replication information was available ) . We used PCR amplification and Sanger-sequencing , identifying SNPs using both automated prediction and manual curation as previously described ( http://pga . gs . washington . edu/ ) . 3′UTRs were defined using the appropriate gene models , while promoters were defined as the 1 kb segment upstream of the annotated transcriptional start site . We subsequently defined haplotypes within each promoter and 3′UTR as previously described using Phase [58] , and designated as common all haplotypes present in at least two samples . Common haplotypes for each of 14 promoter and UTR regions were PCR-amplified and cloned into luciferase-reporter vectors . Promoter haplotypes were cloned immediately upstream of the luciferase reporter gene , while 3′UTRs were placed at the 3′ end of a luciferase gene whose expression is driven by the RPL10 promoter that has strong constitutive activity ( vector maps available from SwitchGear Genomics , http://switchgeargenomics . com/resources/vector-maps/ ) . We then transfected each of these constructs into HEPG2 cells , a liver-derived cell line , and measured luminescence . Each haplotype was tested using multiple ( mode = 3 ) vector preparations and 4 technical transfection replicates measurements were obtained for each vector preparation ( 12 or more measurements for most haplotypes ) . Transient transfection reporter assays were all performed in 96-well format . Transfection complexes were formed by incubating 100 ng of each individual promoter construct with 0 . 3 µL of Fugene 6 transfection reagent and Opti-MEM media in a total volume of 5 µL and incubated for 30 min . Transfection complexes ( 5 uL ) were added to 10 , 000 HepG2 cells in 96-well format that had been seeded 24 h prior to transfection in a white tissue-culture treated plate . After seeding and transfection , cells were incubated for 48 h before freezing at −80 degrees overnight . To read luminescent activity , plates were thawed for 45 min at room temperature . Then 100 µL of Steady-Glo reagent ( Promega #E2520 ) was added and incubated for 30 min at room temperature . Then luminescence was read for 2 s per well on a 384-well compatible plate luminometer ( Molecular Devices LMax384 ) . To identify significant in vitro effects of haplotype on luminescence , we employed a mixed-effects model using the lmer package [63] within R [62] , grouping the replicate luminescence measurements by mini-prep identifier ( treating the mini-prep as a random effect ) . The haplotype identifier has a significant effect on luminescence at p-value<0 . 001 for each of the three reported associations between haplotype sequence and luminescence measurement . No additional correlations were significant at this threshold .
|
Many disease-associated genetic variants do not alter protein sequences and are difficult to precisely identify . Discovery of expression quantitative trait loci ( eQTL ) , or correlations between genetic variants and gene expression levels , offers one means of addressing this challenge . However , eQTL studies in primary cells have several shortcomings . In particular , their reproducibility is largely unknown , the variables that generate unreliable associations are uncharacterized , and the resolution of their findings is constrained by linkage disequilibrium . We performed a three-way replication study of eQTLs in primary human livers . We demonstrated that ∼67% of cis-eQTL associations are replicated in an independent study and that known polymorphisms overlapping expression probes , SNP-to-gene distance , and unmeasured confounding variables all influence the replication rate . We fine-mapped 14 eQTLs and identified causative polymorphisms in the promoter or 3′UTR for 3 genes , suggesting that a considerable fraction of eQTLs are driven by proximal variants that are amenable to functional isolation . Finally , we found hundreds of overlaps between SNPs associated with complex traits and replicated eQTL SNPs . Our data provide both cautionary ( i . e . non-reproducibility of many strong eQTLs ) and optimistic ( i . e . precise identification of functional non-coding variants ) forecasts for future eQTL analyses and the complex traits that they influence .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"genome-wide",
"association",
"studies",
"genome",
"expression",
"analysis",
"genomics",
"gene",
"expression",
"genetics",
"biology",
"human",
"genetics",
"genetics",
"and",
"genomics"
] |
2011
|
Identification, Replication, and Functional Fine-Mapping of
Expression Quantitative Trait Loci in Primary Human Liver Tissue
|
Trypanosoma cruzi is the causative agent of the life-threatening Chagas disease , in which increased platelet aggregation related to myocarditis is observed . Platelet-activating factor ( PAF ) is a potent intercellular lipid mediator and second messenger that exerts its activity through a PAF-specific receptor ( PAFR ) . Previous data from our group suggested that T . cruzi synthesizes a phospholipid with PAF-like activity . The structure of T . cruzi PAF-like molecule , however , remains elusive . Here , we have purified and structurally characterized the putative T . cruzi PAF-like molecule by electrospray ionization-tandem mass spectrometry ( ESI-MS/MS ) . Our ESI-MS/MS data demonstrated that the T . cruzi PAF-like molecule is actually a lysophosphatidylcholine ( LPC ) , namely sn-1 C18:1 ( delta 9 ) -LPC . Similar to PAF , the platelet-aggregating activity of C18:1-LPC was abrogated by the PAFR antagonist , WEB 2086 . Other major LPC species , i . e . , C16:0- , C18:0- , and C18:2-LPC , were also characterized in all T . cruzi stages . These LPC species , however , failed to induce platelet aggregation . Quantification of T . cruzi LPC species by ESI-MS revealed that intracellular amastigote and trypomastigote forms have much higher levels of C18:1-LPC than epimastigote and metacyclic trypomastigote forms . C18:1-LPC was also found to be secreted by the parasite in extracellular vesicles ( EV ) and an EV-free fraction . A three-dimensional model of PAFR was constructed and a molecular docking study was performed to predict the interactions between the PAFR model and PAF , and each LPC species . Molecular docking data suggested that , contrary to other LPC species analyzed , C18:1-LPC is predicted to interact with the PAFR model in a fashion similar to PAF . Taken together , our data indicate that T . cruzi synthesizes a bioactive C18:1-LPC , which aggregates platelets via PAFR . We propose that C18:1-LPC might be an important lipid mediator in the progression of Chagas disease and its biosynthesis could eventually be exploited as a potential target for new therapeutic interventions .
Trypanosoma cruzi is the etiological agent of Chagas disease , which is associated with myocarditis and vasculitis , accompanied by an increase in inflammatory mediators , such as cytokines , chemokines , phospholipids , and glycolipids [1]–[3] . There is also an increase in platelet aggregation , focal ischemia , and myonecrosis in both acute and chronic stages of the disease [2] , [4] , [5] . Most of the chronic cases are linked with debilitating cardiomyopathy , which is responsible for more deaths than any other parasitic disease in Latin America [1] , [6] . Endemic Chagas disease affects eight to ten million people in 21 countries in Latin America [7] . Chagas disease is also becoming a global health problem because of the migration of unconsciously T . cruzi-infected individuals from Latin American countries to other regions of the world . Several thousands of people in the United States , Canada , various European countries , Australia , and Japan are chronically infected with T . cruzi [7] . T . cruzi has a complex life cycle , with two morphophysiological stages within a triatomine bug and two in a mammalian host . Infectious metacyclic trypomastigotes can be expelled with the insect's excreta during a bloodmeal , reaching the host bloodstream through the bite wound or exposed ocular or oral mucosa . In addition , the insect can acquire the parasite during blood feeding from an infected individual and continue the cycle [8] . Blood transfusion , organ transplantation , congenital transmission and food and fluid contamination are other significant ways of transmitting this disease [9] , [10] . In general , lipid mediators [11]–[16] and specifically , lysophosphatidylcholine ( LPC ) [17] , [18] , have been implicated in experimental models of Chagas disease . LPC is present in the saliva of at least one of the insect vectors of Chagas disease , the hemipteran Rhodnius prolixus , where it acts as an anti-hemostatic molecule and immunomodulator of T . cruzi infection in a mammalian model [17]–[19] . LPC ( 1-acyl-2-hydroxy-sn-glycero-3-phosphorylcholine ) is a major plasma phospholipid of oxidized low-density lipoproteins ( Ox-LDL ) , albumin and other carrier proteins , being a critical factor in the inflammatory processes and the atherogenic activity of Ox-LDL [20]–[22] . LPC is an intracellular modulator that activates several second messengers , controlling important biological activities , such as cellular proliferation and differentiation , transcription of adhesion molecules and growth factors in endothelial cells , as well as the transportation of fatty acids , choline , and phosphatidylglycerol between tissues [20]–[24] . The biological activities of LPC are usually mediated by G protein-coupled receptors ( GPCRs ) , such as G2A , GPR4 , and the receptors for prostacyclin ( IP ) , thromboxane A2 ( TXA2 ) ( TP ) , and platelet-activating factor ( PAF ) ( PAFR ) [24]–[33] . Specifically , LPC species are capable of eliciting different cellular activities depending on the length and degree of unsaturation of its sole acyl-chain [30] , [34] , [35] . Trypanosomatid parasites ( e . g . , T . brucei , Leishmania spp . , and T . cruzi ) are known to synthesize phosphatidylcholine ( PC ) and LPC . Over 50% of the total lipids shed to the culture medium by T . cruzi were identified as PC and LPC [36] . These molecules were also found in Leishmania [37] , [38] , African trypanosomes [39] , and in the malaria parasite , Plasmodium falciparum [40] . To the best of our knowledge , however , the chemical structures of LPC species synthesized by T . cruzi have not been defined to date . Platelet-activating factor ( 1-O-alkyl-2-acetyl-sn-glycero-3-phosphocholine; PAF ) is structurally very similar to LPC [41] . PAF exhibits potent biological activity and is synthesized by a wide variety of cells , including neutrophils , platelets , macrophages , and lymphocytes [42] . PAF induces numerous physiological and pathophysiological effects , such as cellular differentiation , inflammation , and allergy , through the activation of specific GPCRs with seven transmembrane helices [25] , [43] . We have previously shown that T . cruzi synthesizes a lipid with platelet-aggregating properties similar to PAF [14] . Preliminary structural analysis by chemical and enzymatic treatment indicated that the T . cruzi PAF-like lipid , metabolically labeled with 14C-acetate , was labile to mild-alkaline or hydrofluoric acid hydrolysis , suggesting a molecule containing a glycerolipid moiety with at least one acyl chain and a phosphate group [14] . However , the detailed structure of the T . cruzi PAF-like lipid remains elusive . Here , we describe the identification and bioactivity of the as-yet elusive T . cruzi PAF-like molecule . We use a novel approach for the enrichment of this and other closely related lysophospholipids , followed by tandem mass spectrometry ( MSn ) to provide ample structural information . Moreover , we constructed a 3-D molecular model of PAFR and used molecular docking to predict the interactions of the T . cruzi PAF-like molecule and other lysophospholipids with the receptor .
Rabbit platelets used in this study were obtained following the guidelines of the Committee for Evaluation of Animal Use for Research of the Federal University of Rio de Janeiro ( CAUAP-UFRJ ) and the NIH Guide for the Care and Use of Laboratory Animals . The vertebrate animal protocol was approved by CAUAP-UFRJ under registry number IBQM011 . Synthetic C16:0- , C18:0- , C18:1 ( Δ9 ) - , and C22:6-LPC , and C16:0-PAF were purchased from Avanti Polar Lipids ( Alabaster , AL ) . The competitive PAF antagonist WEB 2086 ( 4-[3-[4- ( 2-chlorophenyl ) -9-methyl-6h-thieno[3 , 2-f][1] , [2] , [4]triazolo[4 , 3-a]diazepin-2-yl]-1-oxopropyl]morpholine ) was kindly provided by Dr . H . Heurer from Boehringer Ingelheim ( Ingelheim , Germany ) . Otherwise indicated , all other reagents and solvents used here were of analytical , HPLC , or mass spectrometric grade from Sigma-Aldrich ( St . Louis , MO ) . All T . cruzi life cycle stages or forms were obtained from the Y strain [44] . Epimastigote forms ( Epis ) were maintained by weekly transfers using liver infusion tryptose ( LIT ) medium [45] , supplemented with 0 . 002% hemin and 10% heat-inactivated fetal calf serum ( FCS; Hyclone , heat-inactivated at 56°C for 30 min ) at 28°C . Metacyclic trypomastigote forms ( Metas ) were obtained by spontaneous axenic differentiation of Epis at 28°C , followed by their purification using ion-exchange chromatography [46] , [47] . Mammalian tissue culture-derived trypomastigotes ( TCTs ) were obtained from the supernatants of 5 to 6 days old T . cruzi-infected LLC-MK2 cells ( American Type Culture Collection , Rockville , MD ) , maintained in RPMI-1640 medium supplemented with 2% FCS at 37°C in a 5% humidified CO2 atmosphere [48] . Intracellular amastigotes ( ICAs ) were obtained as described [49] . Briefly , infected monolayers of LLC-MK2 cells were gently detached by scraping ( BD Falcon cell scraper , BD Biosciences ) and resuspended in PBS supplemented with 10% FCS . Mammalian cells were disrupted by passage through a 27-gauge needle ( BD , Becton and Dickinson & Co . ) . ICA forms were separated from the cell debris by centrifugation ( 800× g for 5 min at 4°C ) . The supernatant was then harvested and passed through a DE-52 column and parasites were incubated for 2 h at 37°C in a humidified 5% CO2 atmosphere , after which the parasites were again passed through a DE-52 column , from which they were harvested and stored . For the viability testing of all parasite forms , cells were resuspended in a Trypan Blue solution and counted in a Neubauer chamber [50] . In this study , all experiments were performed using parasites that were harvested by centrifugation and washed three times with PBS before use , unless otherwise specified . All parasite forms were counted and then frozen in liquid nitrogen prior to use . Frozen pellets derived from Epis , Metas , ICAs , and TCTs ( 2×109 cells each ) , were suspended in 1 . 6-ml ice-cold HPLC-grade water and transferred to 13×100-mm Pyrex culture tubes with polytetrafluoroethylene ( PTFE ) -lined screw caps . HPLC-grade chloroform and methanol were added to each vial , giving a final ratio of chloroform/methanol/water ( C/M/W ) of 1∶2∶0 . 8 ( v/v/v ) . The samples were mixed vigorously using a vortex for 2 min and then centrifuged for 15 min at 1 , 800× g at room temperature . After centrifugation , the supernatants were transferred to PTFE-lined Pyrex glass test tubes and the pellets were dried under a constant flow of N2 stream . The dry pellets were then extracted three times with C/M ( 2∶1 , v/v ) and twice with C/M/W ( 1∶2∶0 . 8 , v/v/v ) . After extraction , the supernatants were pooled together and dried before being subjected to Folch's partition [51] . To this end , samples were first dissolved in C/M/W ( 4∶2∶1 . 5 , v/v/v ) and then mixed vigorously for 5 min using a vortex and finally centrifuged for 15 min at 1 , 800× g at room temperature . After centrifugation , the lower ( organic ) and upper ( aqueous ) phases were separated in PTFE-lined Pyrex glass test tubes . The Folch lower phase was then washed two times with a freshly prepared upper phase , dried under N2 steam , and stored at −70°C until use . Phospholipids derived from the lower phase of the Folch's partition , as described above , were purified from other classes of lipids using a three-step SPE protocol [52] . Briefly , 100 mg silica gel ( Merck , grade 7754 , high purity , 70–230 mesh , 60 Å ) were packed into borosilicate glass Pasteur pipettes ( 5 ¾″ , Fisher Scientific ) using Pyrex glass fiber wool ( 8-µm pore size , Sigma-Aldrich ) as a sieve . The column was sequentially conditioned with 4 ml each methanol , acetone , and chloroform . Dried Folch lower phase samples from all T . cruzi life stages were redisolved in 3 ml chloroform and a third of each sample was added to the column . Lipids were sequentially eluted with 4 ml chloroform ( neutral lipids ) , acetone ( glycolipids ) , and methanol ( phospholipids and free fatty acids ) . Each fraction was collected into a 7-ml amber glass vial with PTFE-lined screw top ( SUPELCO , Sigma-Aldrich ) . All samples were immediately dried under a constant flow of N2 stream and stored at −70°C until use . With the purpose of enriching the putative PAF-like molecule from a complex T . cruzi phospholipid mixture , a novel method was developed using perfusion chromatography [53] . Briefly , fifty microliters of a suspension of 40 mg/ml POROS R1 50 ( poly[styrene/divinylbenzene] , with similar binding strength as C4 supports ) beads ( Applied Biosystems , 50-µm diameter ) in HPLC-grade n-propanol ( Honeywell , Burdick & Jackson , Radnor , PA ) were packed into a 200-µl sterile micropipette tip ( Axygen , Corning Life Sciences , Union City , CA ) . Pyrex glass fiber wool ( 8-µm pore size , Sigma-Aldrich ) was used as sieve . The POROS R1 mini-column was washed twice with 100 µl HPLC-grade methanol and then conditioned with an n-propanol/water gradient ( from 50% to 0% , in 5% increments ) . Each step of the gradient was performed with 100 µl solvent , with the exception of the last step ( 0% n-propanol ) , which was performed twice with 100 µl HPLC-grade water . Then , either the mixture of lipid standards ( 100 pmol C16:0-lyso-PC ( C16:0-LPC ) , C16:0-lyso-PAF ( C16:0-LPAF ) , C16:0-PAF , and C18:0/C18:1-diacyl-PC ) or T . cruzi phospholipids derived from the SPE procedure were suspended in HPLC-grade water , sonicated for 10 min in a bath sonicator , and added to the column . Both the standards and the phospholipid extracts were eluted using a 0%–50% n-propanol gradient in 5% n-propanol increments . Each fraction was stored in Axygen 2-ml microcentrifuge tube at −70°C until further use . All fractions derived from the POROS R1 mini-column purification , as well as the fractions obtained prior to this last procedure ( namely , the lower Folch and methanol phase ) , were dissolved in MS-grade methanol containing 5 mM LiOH , as indicated . Samples were directly injected ( at 300 nl/min ) by chip-based infusion using a TriVersa NanoMate nanoelectrospray source ( Advion , Ithaca , NY ) , into an LTQXL ESI-linear ion trap-MS ( ESI-LIT-MS ) ( Thermo Fisher Scientific ) , in positive-ion mode . The source voltage was set at 0 . 01 kV and current at 0 . 03 µA; capillary voltage and temperature were 36 V and 150°C , respectively; and tube lens voltage was set at 145 V . Select ions were subjected to sequential tandem fragmentation ( MSn ) by collision-induced dissociation ( CID ) . Full-scan ( MS ) spectra were collected at the 400–1000 m/z range . Tandem mass fragmentation was carried out using normalized collision energies of 35 , 40 , and 45 for MS2 , MS3 , and MS4 , respectively . The resulting spectra were compared to the aforementioned phospholipid standards , as well as to previously described results . The location of the acyl chain on T . cruzi LPC species was determined by MS2 and M3 analysis , essentially as described by Hsu et al . [54] . Briefly , sn-1 and sn-2 C18:1-LPC regioisomer standards were generated by treatment of 18:1 ( Δ9-cis ) -PC ( 1 , 2-dioleoyl-sn-glycero-3-phosphocholine , catalog # 850375 , Avanti Polar Lipids ) with either PLA2 ( from porcine pancreas , catalog # P6534 , Sigma-Aldrich ) or PLA1 ( from Thermomyces lanuginosus , catalog # L3295 , Sigma-Aldrich ) . For PLA1 treatment , one mg of diacyl-PC standards were dried under N2 stream , redisolved in 200 µL of reaction buffer ( 50 mM Tris-HCl , 2 mM CaCl2 , 140 mM NaCl , pH 8 . 0 ) and sonicated for 30 min . Afterwards , the samples were incubated at 37°C for 2 h in the presence of 12 units PLA1 . The reaction was interrupted by the addition of 1 . 5 mL chloroform , followed by vortexing for 1 min . The resulting LPC species present in the organic phase of the mixtures were then purified by SPE , following the protocol described above . Finally , LPCs were recovered in the methanolic phase of the SPE column . PLA2 treatment was performed following the same procedure steps used for PLA1 treatment but using a different reaction buffer ( 100 mM Tris-HCl , 5 mM CaCl2 , 100 mM NaCl , pH 8 . 0 ) and 5 units PLA2 [55] . The purified lipids ( methanolic phase of the SPE ) were analyzed by ESI-LIT-MS in 100% methanol containing 5 mM LiOH or 5 mM NaCl . The MS analyses were performed as described above . The assignment of the fatty acid position on the LPC ( sn-1 or sn-2 ) was performed by comparing the fragmentation pattern of standard sn-1 and sn-2 C18:1-LPCs with T . cruzi-derived 18:1-LPC samples . The position of the double bond on fatty acid substituents was determined by MS4 , essentially as described by Hsu et al . [56] . Briefly , sn-1 C18:1 ( Δ6-cis ) -LPC and sn-1 C18:1 ( Δ9-cis ) -LPC standards were generated by treatment of both 18:1 ( Δ6-cis ) -PC ( 1 , 2-dipetroselenoyl-sn-glycero-3-phosphocholine , catalog # 850374 , Avanti Polar Lipids ) and 18:1 ( Δ9-cis ) -PC ( 1 , 2-dioleoyl-sn-glycero-3-phosphocholine , catalog # 850375 , Avanti Polar Lipids ) with PLA2 ( from porcine pancreas , catalog # P6534 , Sigma-Aldrich ) , as described above . LPC species were recovered from incubation mixtures and analyzed by ESI-LIT-MS in 100% methanol containing 5 mM LiOH , as described above . Commercial C10:0-LPC ( 1-decanoyl-2-hydroxy-sn-glycero-3-phosphocholine , catalog # 855375 , Avanti Polar Lipids ) was used as an internal standard to quantify the most abundant LPC species found in T . cruzi samples . Briefly , 18 nmoles of standard C10:0-LPC were added to parasite pellets ( 7×108 cells ) shortly before lipid extraction . Lipid extraction was conducted on freshly prepared parasite pellets following the protocol described above . Then , the Folch lower-phase fractions were analyzed by ESI-MS ( at the 400–1000 m/z range ) under the identical MS conditions used for the characterization of T . cruzi LPCs . The amount of each LPC species was calculated using the formula: [T . cruzi LPC peak intensity/C10:0-LPC peak intensity]×[concentration of C10:0-LPC ( 18 pmol/µl ) /MRRF] , where , MRRF stands for the molar relative response factor of each LPC species to the C10:0-LPC standard . The MRRF was calculated by dividing the intensity of the peak corresponding to a standard LPC ( C16:0-LPC , C18:0-LPC , or C18:1-LPC ) by the intensity of the peak corresponding to C10:0-LPC ( m/z 418 . 5 ) , when all molecules were in equimolar concentrations . T . cruzi LPC species were also quantified in extracellular vesicles ( EV ) and EV-free supernatant or conditioned medium ( VF ) of Epis ( eV2 , eV16 , and eVF ) and Metas ( mV2 , mV16 , and mVF ) , obtained from Epi and Meta pellets ( 9×109 parasites each ) , as previously described [57] . C10:0-LPC ( m/z 418 ) ( 18 nmoles per sample ) was used as an internal standard . LPC species were extracted from Epi- and Meta-derived EVs , and respective total parasite pellets ( ePellet and mPellet ) as described above . The Folch lower phase fractions were analyzed by ESI-LIT-MS as above . C16:0-PAF and various synthetic ( C16:0- , C18:0- , C18:1- , and C22:6-LPC ) or purified ( C18:2-LPC ) LPC species were tested in a platelet aggregation assay [58] . Rabbit blood platelets were prepared from blood collected with 5 mM EDTA as anticoagulant , isolated by centrifugation , washed and resuspended in a modified Tyrode's buffer , pH 7 . 4 , containing 2 mM CaCl2 , at a final concentration of 3–4×105 cells/µl in Tyrode's buffer . Platelet aggregation experiments were performed with a Chronolog Aggregometer ( Havertown , PA , USA ) , with monitoring time of 5 min . Rabbit platelets used in this investigation were obtained following the guidelines for animal experimentation of the USA National Institutes of Health and the experimental protocol received official approval of the Institutional Animal Care and Use Committee , Universidade Federal do Rio de Janeiro . The amino acid sequence of the human PAF receptor ( PAFR , UniProtKB ID: P25105 ) was obtained from ExPASy server [59] . The region between Asp10-Ser310 , part of PAFR sequence that includes all seven-transmembrane domains , was submitted to I-TASSER server , which combines threading and ab initio algorithms [60] , [61] . The I-TASSER server , ranked as the best server in recent CASP7 and CASP8 experiments , builds protein models based on multiple-threading alignments by LOMETS program and iterative TASSER program simulations [61] , [62] . In addition , MODELLER v9 . 10 program [63] , [64] ( http://salilab . org/modeller/ ) was used to add a disulfide bridge between Cys90-Cys173 and , subsequently , to refine the best I-TASSER model . Thus , the final model was validated using three programs: PROCHECK [65] and ERRAT [66] , both at SAVES server ( http://nihserver . mbi . ucla . edu/SAVES_3/ ) and PROQM [67] ( available as a server at http://www . bioinfo . ifm . liu . se/services/ProQM/index . php ? about=proqm ) . PROCHECK analyzes the stereochemical quality and ERRAT evaluates the non-bonded atomic interactions in the model structure , while PROQM uses a specific-scoring function for membrane protein , including GPCR , to assess local and global structural quality of the model . The ligand structures ( C16:0-PAF , C16:0-LPC , C18:0-LPC , C18:1-LPC , and C18:2-LPC ) were built in the Spartan'10 software ( Wavefunction , Inc . , Irvine , CA ) . The docking of the ligands to the PAFR model binding site was performed using Molegro Virtual Docker ( MVD ) program ( CLC bio , Aarhus , Denmark ) , which uses a heuristic search algorithm that combines differential evolution with a cavity prediction algorithm . The MolDock scoring function used is based on a modified piecewise linear potential ( PLP ) with new hydrogen bonding and electrostatic terms included . Full description of the algorithm and its reliability compared to other common docking algorithm have been described [68] . As no satisfactory cavities were found by cavity prediction algorithm using MVD , His248 ( a constituent residue of the binding pocket ) was set as center of searching space . The search algorithm MolDock optimizer was used with a minimum of 50 runs and the parameter settings were: population size = 200; maximum iteration = 2000; scaling factor = 0 . 50; offspring scheme = scheme 1; termination scheme = variance-based; crossover rate = 0 . 90 . Due to the stochastic nature of algorithm search , ten independent simulations per ligand were performed to predict the binding mode . Consequently , the complexes with the lowest interaction energy were evaluated . The interactions between PAFR and each ligand were analyzed using the ligand map algorithm , a standard algorithm in MVD program . The usual threshold values for hydrogen bonds and steric interactions were used . All figures of PAFR modeling and docking were edited using Visual Molecular Dynamics ( VMD ) program ( available for download at http://www . ks . uiuc . edu/Research/vmd/vmd-1 . 9 . 1/ ) .
Previous results from our group strongly indicated that T . cruzi synthesizes a phospholipid with platelet-aggregating activity similar to PAF [14] . Thus far , however , the precise structure of this bioactive parasite-derived molecule remains unknown . Aiming at the enrichment and characterization of the putative T . cruzi PAF-like phospholipid from a complex phospholipid mixture , we developed a fractionation protocol , which included solvent extraction and Folch's partition , followed by SPE and perfusion chromatography ( Fig . 1 ) . Lipid fractions obtained after each step of purification were analyzed by ESI-LIT-MS in positive-ion mode . Total-ion mapping ( TIM ) for the neutral loss of the trimethylamine group ( = 59 a . m . u . ) was performed to promptly localize phosphocholine-containing phospholipids of all life-cycle stages of T . cruzi ( data not shown ) . ESI-LIT-MS analysis of the Folch lower phase of the four parasite forms ( Epi , Meta , ICA , and TCT ) showed major phospholipid species at the 700–900 m/z range , except for the ICA form ( Fig . 2A ) . Tandem MS ( MSn ) analysis of these lipid species revealed that , as expected , they were mostly diacyl-PC and sphingomyelin ( SM ) species ( data not shown , to be published elsewhere ) . In contrast to other parasite forms , ICA is much richer in lipid species at the 500–600 m/z range , particularly m/z 526 , 528 , 530 , and 574 ( Fig . 2A ) . Noteworthy , lithiated singly-charged ion species ( [M+Li]+ ) of synthetic LPAF , LPC , and PAF standards were found at this range of the spectrum ( Fig . S1 , top spectrum ) . ESI-LIT-MS analysis of SPE-derived fractions of all T . cruzi forms revealed a clear enrichment of phosphocholine-containing lipids at the low m/z range ( 400–600 ) , which would indicate an enrichment of potential LPC , LPAF , or PAF molecules . Nevertheless , most samples still contained high amounts of diacyl-PCs and , possibly , SMs ( data not shown ) . Therefore , a novel protocol using POROS R1 beads was designed to enrich the putative PAF-like molecule , which as we predicted could have a structure similar to PAF , LPAF , or LPC . First , we tested the POROS R1 mini-column with a complex mixture of phospholipid standards containing LPAF , LPC , PAF , and diacyl-PCs . We were able to obtain highly enriched LPAF , LPC , and PAF species in the 20% and 25% n-propanol fractions ( Fig . S1 ) . Identical conditions were applied for further fractionation of the phospholipids present in the SPE methanolic fraction of all T . cruzi forms . The 25%-n-propanol fractions from these parasite stages were then compared by positive-ion mode ESI-LIT-MS , using the same concentration of cells ( 4×105/µl ) and flow rate ( 300 nl/min ) ( Fig . 2B ) . The ion species at the 700–900 m/z range , corresponding to diacyl-PCs and SMs , were noticeably much less abundant in the 25%-n-propanol POROS R1 fraction ( Fig . 2B ) than in the Folch lower phase and SPE methanolic fractions ( Fig . 2A and data not shown ) , which is in agreement with the observed phospholipid standard results ( Fig . S1 ) . In contrast , the ion species at the 400–600 m/z range were prominently more abundant in the 25%-n-propanol POROS R1 fraction than in lower Folch and SPE methanolic fractions ( Fig . 2A , B and data not shown ) . In particular , the ion species at m/z 526 and 528 were remarkably abundant in the infective TCT form and in the noninfective ICA form . Tandem MS was then performed for the elucidation of the molecular structures of all phosphocholine-containing lysophospholipids at the 400–600 m/z range . The fragmentation pattern of the synthetic C16:0-PAF standard ( m/z 530 ) was compared to those of synthetic C18:0- and C18:1-LPC standards ( m/z 530 and 528 , respectively ) , because certain PAF and LPC species may be isobaric . The assignment of the lysophospholipid species found in the 25%-n-propanol POROS R1 fraction of all T . cruzi forms was based on the fragmentation pattern of these standards , as well as on previously reported results [41] , [54] , [56] . Tandem MS ( MS2 ) analysis of singly-charged , lithiated C16:0-PAF , C18:0-LPC , and C18:1-LPC ion species gave rise to fragment ions at m/z 471 , 471 , and 469 , respectively , corresponding to the neutral loss of 59 a . m . u . ( = trimethylamine group ) . The fragmentation of 16:0-PAF standard , however , also gave rise to a fragment ion at m/z 341 , consistent with the neutral loss of the whole phosphocholine headgroup along with the lithium adduct ( −189 a . m . u . ) . This fragment could not be detected on either LPC standards ( Fig . S2A ) . MS3 Fragmentation of the major ions obtained by MS2 of C16:0-PAF , C18:0-LPC , and C18:1-LPC ( i . e . , m/z 471 , 471 , and 469 ) , gave rise to the major non-lithiated ion fragments at m/z 341 , 341 , and 339 , respectively , corresponding to the loss of 130 a . m . u . ( = ethyl phosphate group+Li ) ( Fig . S2B ) . In addition , we observed lithiated ion fragments at m/z 347 , 347 , and 345 , corresponding to the loss of 124 a . m . u . ( = ethyl phosphate group ) , for C16:0-PAF , C18:0-LPC , and C18:1-LPC , respectively . Interestingly , C18:0-LPC and C18:1-LPC also gave rise to two fragment ions ( m/z 291 and 289 , respectively ) that could not be detected in the C16:0-PAF standard . These fragment ions corresponded to the lithiated ( [R1CO2H+Li]+ ) stearoyl ( m/z 291 ) and oleyl chains ( m/z 289 ) , after the loss of choline ( N+ ( CH3 ) 3 ( CH2 ) 2OH ) from the precursor ions m/z 427 and 425 , respectively ( Fig . S2B ) [54] . Finally , we carried out MS4 analysis of the major fragment ion species obtained by MS3 of C16:0-PAF , C18:0-LPC , and C18:1-LPC standards ( m/z 341 , 341 , and 339 , respectively ) ( Fig . 3 ) . A complex fragmentation pattern that provided ample structural information for all three standards was observed . In these spectra , it was possible to identify fragment ions generated by the loss of the acetyl group at the sn-2 position of C16:0-PAF ( m/z 281 and 263 ) , and a fragment ion corresponding to the protonated C16:0-alkyl chain ( R1+ ) ( m/z 225 ) . Moreover , a series of fragments resulting from the loss of methylene groups ( = 14 a . m . u . ) from the PAF C16:0-alkyl chain could also be seen below m/z 225 . Similar information could be obtained from the C18:0-LPC and C18:1-LPC , where fragment ions derived from the protonated stearoyl ( m/z 285 , 267 , and 249 ) and oleyl ( m/z 265 and 247 ) acyl chains at the sn-1 position could be identified [54] . In both cases , a series of fragments corresponding to the loss of methylene units could also be seen below m/z 240 ( Fig . 3 ) . Based on the fragmentation pattern of the standards in MS2 , MS3 , and MS4 , we could assign the different lysophospholipid species of T . cruzi enriched in the POROS R1 25%-n-propanol fraction . As observed in Figs . 3 and S2A , B , the parent ions at m/z 526 , 528 , and 530 corresponded to C18:2-LPC , C18:1-LPC , and C18:0-LPC , respectively . The lithiated ( [R1CO2H+Li]+ ) , non-lithiated ( [R1CO+]+ , and dehydrated [R1CO+ - H2O]+ ) fragment ions of linoleyl ( m/z 287 , 263 , and 245 ) , oleyl ( m/z 289 , 265 , and 247 ) , and stearoyl ( m/z 291 , 285 , 267 , and 249 ) chains , respectively , corroborated our assignments of T . cruzi ( Tc ) m/z 526 , 528 , and 530 as C18:2- , C18:1 , and C18:0-LPC ( Figs . 3 and S2A , B ) . No detectable traces of C16:0-PAF ( isobaric to C18:0 LPC ) or any other PAF-like species could be found . The structure and the fragmentation pattern of C16:0-PAF and the major T . cruzi LPC species are represented in Fig . 4 . The MSn experiments carried out above , however , could not provide sufficient structural information to determine the position of the fatty acid ( sn-1 or sn-2 ) and the location of the double bonds on three major LPCs of T . cruzi . To address this point , we first generated LPC standards with fatty acids on either the sn-1 or sn-2 position by treating diacyl-PCs with commercial PLA2 and PLA1 , respectively . Following protocols by Hsu et al . [54] , we were able to determine that the acyl chain in the three major T . cruzi LPCs was localized at the sn-1 position . In Figs . S3A , B , the fragmentation spectra ( MS2 and MS3 ) of sodiated and lithiated sn-1 C18:1-LPC , sn-2 C18:1-LPC , and T . cruzi C18:1-LPC ( from ICA form ) are shown . In agreement with Hsu et al . [54] , the relative abundance of the sodiated parent ion ( m/z 544 ) to the fragment ion ( m/z 485 ) , corresponding to the loss of trimethylamine ( −59 a . m . u ) , could be used to differentiate between the two possible regioisomers ( Fig . S3A ) . Clearly , T . cruzi C18:1-LPC showed a fragmentation pattern consistent with an acyl chain located at sn-1 . This result was corroborated by the fragmentation spectrum of the lithiated T . cruzi C18:1-LPC ( Fig . S3B ) . In this case , the relative abundance of the fragment ion at m/z 425 [M – N+ ( CH3 ) 3 ( CH2 ) 2OH+Li+]+ to the ions at m/z 339 ( M - 189 ) and m/z 345 ( M - 183 ) was used to corroborate the sn-1 position of the acyl chain on T . cruzi C18:1-LPC . We carried out identical experiments with T . cruzi C18:0- and C18:2-LPC and found that both contained the acyl chain at the sn-1 position ( data not shown ) . To address the location of the double bonds in T . cruzi C18:1- and C18:2-LPC , we followed the protocols described by Hsu and Turk [69] . By comparing the MS4 spectra of Δ9- and Δ6-C18:1-LPC standards , we observed noticeably different fragmentation patterns of the acyl chain , especially in the relative abundance of fragment ions C16H31 ( m/z 223 ) , C14H29 ( m/z 197 ) , C14H27 ( m/z 195 ) , C13H27 ( m/z 183 ) , C13H25 ( m/z 181 ) , C8H15 ( m/z 111 ) , and C8H13 ( m/z 109 ) ( Fig . S3C ) . When T . cruzi C18:1-LPC ( from ICA forms ) was analyzed under the same MS conditions , the fragmentation pattern observed was consistent with a Δ9 double bond ( Fig . S3C , bottom spectrum ) . The same type of experiment was conducted with T . cruzi C18:2-LPC ( from ICA forms ) and the resulting fragmentation was consistent with Δ9 , 12 double bonds ( data not shown ) . The POROS R1 protocol we have described here also enriched other LPC species , which included C22:6-LPC ( m/z 574 ) , C22:4-LPC ( m/z 578 ) , C16:0-LPC ( m/z 502 ) , and C16:1-LPC ( m/z 500 ) that were also characterized by MSn ( Fig . S4 ) . Most of these species , except for C22:6-LPC , had very low abundance and , in the case of Epi and Meta forms , could only be seen in the enriched 25%-n-propanol POROS R1 fraction . Even for TCT and ICA forms , MS3 and MS4 of the C16:1- , C22:4- and C18:0-LPC species could only be conducted with samples derived from the POROS R1 chromatography . This confirms that indeed this last fractionation step is necessary for the full characterization of low-abundance LPC species from complex phospholipid mixtures of T . cruzi . Using the current methodology , however , we were unable to detect any bona fide PAF species in any of the four parasite stages analyzed . Even employing the highly specific and sensitive MS approach , selective-ion monitoring ( SIM ) , we could not detect any trace amounts of PAF species in T . cruzi . This was true not only for the POROS R1 25% n-propanol fractions , but for all fractions described in this study . Therefore , we surmise that if there were any PAF species in this parasite , the concentration levels would likely be below the detection limit of the MS approaches used in the present study . After characterizing the different T . cruzi LPC species , we proceeded to quantify them using a synthetic C10:0-LPC as an internal standard . These analyses were conducted on freshly prepared parasite pellets to minimize the amount of time that T . cruzi phospholipases A1 and A2 [70] could have to act on PCs , leading therefore to an artificial increase in LPC levels . Overall , the lipid profiles of freshly prepared pellets were nearly identical to the profiles of previously frozen pellets of the same parasite forms ( Figs . 2 and 5 , and data not shown ) . The LPC quantification showed that the amount of C16:0- , C18:0- , C18:1- and C18:2-LPC present in the mammalian-dwelling forms of T . cruzi ( ICA and TCT ) were much higher than in the insect-dwelling ( Epi , Meta ) forms ( Table 1 ) . For instance , although the two most abundant LPC species found in all four parasite forms were C18:1- and C18:2-LPC , ICA forms had ∼44 and ∼35 times more of these species , respectively , than Epis . TCTs also showed very high levels of these lipids , having approx . 16–17 times more C18:1- and C18:2 LPC than Epis ( Table 1 ) . Interestingly , TCTs also contained the highest levels of C16:0-LPC , which has been shown to have immunosuppressant activities in the context of T . cruzi infection [17] . To assess whether T . cruzi is also able to secrete LPC species to the extracellular milieu , we analyzed extracellular vesicles ( EVs ) and EV-free supernatant or conditioned medium ( VF ) of Epi and Meta forms , obtained as described [57] . Epi and Meta forms secrete two types of EVs , namely V2 and V16 , which are vesicles obtained after 2 h and 16 h of ultracentrifugation ( at 100 , 000×g ) , respectively . In both stages , V2 are larger vesicles resembling ectosomes ( 130–140 nm ) , whereas V16 are smaller vesicles resembling exosomes ( 70–90 nm ) . The final vesicle-free supernatant ( VF ) in both stages is virtually devoid of EVs [57] . As shown in Table S1 and Fig . S5 , although Epis have higher amounts of C18:1- and C18:2-LPC than Metas in the total parasite pellet , the latter secrete much higher amounts of these phospholipids in the EV-free conditioned medium ( mVF ) . Metas secrete 1 . 9 pmol of C18:1-LPC per 106 cells in mVF and an additional 0 . 5 pmol associated with mV2 fraction . Similar values were found for C18:2-LPC in mVF and mV2 ( Table S1 ) . In comparison , Epis only secrete trace amounts of both LPCs in eV2 . In our analyses , however , we were only able to detect trace amounts of C16:0-LPC in the V2 , V16 , and VF fractions of Epis and Metas ( data not shown ) . Taken together , our data strongly indicate that C18:1-LPC , C18:2-LPC , and eventually other LPCs could actively be secreted by Metas during the early stages of the infection . Moreover , LPCs secreted by these parasites could contribute to the overall LPC pool in the plasma , saturating the lipid carrier proteins , and eventually activating PAF receptors . Most importantly , the relative concentration of C18:1-LPC could even be higher in the infected tissues , due to the continuous secretion of this phospholipid , associated or not with parasite EVs . Owing to the technical difficulty in obtaining enough parasites for preparation of EVs from TCTs or ICAs , we have not been able to conduct the same kind of quantification in these mammal-dwelling stages . Since we were unable to detect any PAF species in any of the four parasite stages analyzed , we hypothesized that the PAF-like molecule could be one or more of the lysophospholipids structurally characterized here . This assumption is also based on previous reports that certain LPC species seem to activate the PAFR [28] , [33] . Moreover , purification of sufficient amounts of each T . cruzi LPC species for the bioassay ( i . e . , platelet aggregation ) was not feasible . Therefore , we carried out the platelet aggregation assays using synthetic ( C16:0- , C18:0- , C18:1- , and C22:6-LPC , at 1 , 10 , 100 , and 1000 µM ) or purified ( C18:2-LPC ) LPC species , and synthetic C16:0-PAF ( at 1 µM; positive control ) . In a set of experiments platelets were pre-incubated for 30 min in the presence of 10 µM WEB 2086 . As seen in Fig . 6 , C16:0-LPC , C18:0-LPC , and C18:2-LPC failed to aggregate platelets at 10 µM , or at any other concentrations used ( data not shown ) . Interestingly , C18:1-LPC was able to aggregate rabbit platelets at 10 and 100 µM ( Fig . 6 ) , but unlike PAF , failed to perform this activity at 1 µM ( data not shown ) . WEB 2086 completely abolished platelet aggregation induced by C16:0-PAF and C18:1-LPC ( at final concentrations of 1 µM and 100 µM , respectively ) ( Fig . 6 ) . We were aware of the fact that depending upon the acyl chain length and degree of unsaturation , high concentrations ( >30 µM ) of LPCs could act as strong detergent and promote cell lysis [34] . Therefore , in our platelet aggregation assays , we used all LPCs at a maximal 10 µM concentration , except for C18:1-LPC , which was also tested at 100 µM . We did not observe platelet lysis up to 10 µM of any LPC tested , or even at 100 µM of C18:1-LPC . To gain insights into the mechanism by which C18:1-LPC , but not C16:0 , C18:0 , C18:2 , could induce platelet aggregation mediated by PAFR , we decided to build a 3D-structural model of the PAFR , since a high-quality model was not available in the literature . The structural features of our PAFR model are shown in Figure 7A . As a member of the class A of GPCR superfamily , the PAFR model encompasses an extracellular N-terminus , followed by seven transmembrane ( TM ) α-helices ( TM1 , TM2 , TM3 , TM4 , TM5 , TM6 , and TM7 ) ( Fig . S6A ) , connected by three extracellular loops ( EL1 , EL2 , and EL3 ) ( Fig . S6B ) and three intracellular cytoplasmic loops ( CL1 , CL2 , and CL3 ) ( Fig . S6C ) and , finally , a short α-helix ( H8 ) at the intracellular C-terminus ( Fig . 7A ) . Additionally , the spatial arrangement of the 7-TM bundle constitutes a hydrophilic cavity covered by EL2 , which is described as a ligand-binding pocket ( Figs . 7A and S6B ) [68] , [71]–[76] . The I-TASSER methodology is very accurate for the construction of protein models when the sequence identity between the target sequence and the template protein drops below 30% , where lack of a high-quality structure match may provide substantial alignment errors and , consequently , poor quality models . In addition , the I-TASSER methodology has already been used for modeling other GPCRs [68] , [76] . The structural validation of the PAFR model was performed using three programs: PROCHECK [65] , ERRAT [66] , and PROQM [67] . The stereochemical quality of the PAFR model was evaluated using PROCHECK program . According to an analysis of 118 known protein structures , a good quality model would be expected to have over 90% in the most favored regions . The analysis of the first Ramachandran plot revealed that 97 . 8% of the amino acid residues are located in the most favorable ( 92 . 0% ) and additional allowed ( 5 . 8% ) regions , confirming the excellent quality of the PAFR model described here ( Fig . S7 ) . The analysis of the second Ramachandran plot , which considers only Φ and ψ angles for Gly residues , showed that all 11 Gly residues in the primary sequence of PAFR were in allowed regions , with favorable combinations of angles Φ and Ψ ( Fig . S8 ) . The properties of the main and side chains for PAFR model were also evaluated by PROCHECK program , which showed acceptable values for our model when compared to experimentally-determined protein structures ( Fig . S9 and S10 ) . The analysis of distortions in the geometry of the PAFR model residues was also analyzed using PROCHECK ( Fig . S11 ) . The parameters analyzed were the bond lengths and angles , including atoms of the main and side chains , and the results were considered acceptable for our PAFR model . In addition , we used the ERRAT program for verifying errors in non-bonded atom-atom interactions in our model . The error values are plotted as a function of the position of a sliding residue in the window [66] . According to this analysis , structures at high resolutions , generally produce values around 95% or higher . However , for protein structures at lower resolutions ( 2 . 5 to 3 Å ) , the average overall quality factor can be around 91% . The ERRAT analysis gave an overall quality factor value of 87% for our PAFR model ( Fig . S12 ) . Although low for crystal structures , this value is acceptable for protein models [67] . Finally , we analyzed the quality of the three-dimensional structure of our model at local and global levels , using PROQM program [66] ( Fig . S13 ) . This program uses a score function to evaluate structures of membrane proteins , including GPCRs [67] , [77] , [78] . Thus , to each residue of the protein model is given a score , producing , consequently , an overall quality factor . The value of the overall quality factor generated by server PROQM ranges from 0 to 1 , indicating a model of poor and excellent quality , respectively [64] . The PROQM analysis gave a global quality score of 0 . 717 for our PAFR model ( Fig . S13 ) , which is similar to the score values for crystallographic structures of GPCR [67] . To compare the binding mode of all ligands ( C16:0-PAF , and C16:0- , C18:0- , C18:1- , and C18:2-LPC ) to the PAFR model , we carried out ten docking simulations for each ligand ( totalizing 50 poses per ligand ) and , consequently , the poses with the lowest energy for each ligand was selected for analysis . The comparison between C16:0-PAF and each LPC species is represented in Fig . 7B–E and Fig . S14 . The molecular docking study showed that C18:1-LPC and C16:0-PAF have similar modes of interaction with PAFR ( Fig . 7D and Fig . S14C ) . On the other hand , no other LPC species was able to interact with PAFR in a similar way , which corroborates the platelet aggregation assays ( Fig . 6 ) . Since PAF and LPC have different structures , RMSD matrix was not calculated . Instead , distances between the heteroatoms of the functional groups ( amino , phosphate , and acyl groups ) of C16:0-PAF and all LPC species were measured . The nitrogen of the amino group , the phosphorous of the phosphate group , and the oxygen linking the glycerol backbone ( at sn-1 ) to the carbonyl group of the acyl chain were chosen and all distances between the ligands are presented in Table 2 . The distances between heteroatoms of C16:0-PAF and C18:1-LPC are clearly lower than the distances between C16:0-PAF and other LPC species ( C16:0 , C18:0 , and C18:2 ) , with the exception of the amino group of C18:2-LPC and C16:0-PAF , confirming the similarities between C16:0-PAF and C18:1-LPC binding modes . The energy of the interactions between PAFR and each ligand was also analyzed ( Table 3 ) . The hydrogen bonds and repulsive steric interactions were mapped using a ligand-map algorithm , generated by the MVD program [68] . The hydrogen bonds are represented in Figure 8 . C16:0-PAF and C18:1-LPC interact with the same amino acids ( i . e . , Asn159 and Thr160 ) of the PAFR model . Other LPCs do not present the same pattern of interaction . Interestingly , C18:1-LPC is able to additionally interact with Ser157 , through its hydroxyl group , in a very strong way ( Fig . 8 and Table 3 ) . Although C16:0- , C18:0- , and C18:2-LPC also form hydrogen bonds with amino acids , these interactions occur in a different region of the PAFR model and with lower strength . Finally , the amino group either of C16:0-PAF or any of the LPC species did not seem to interact to our PAFR model ( Fig . S15 ) .
Previous results from our group showed that T . cruzi synthesizes a PAF-like phospholipid capable of aggregating rabbit platelets [14] . Aiming at the purification and structural characterization of this putative PAF-like molecule , we developed a fractionation protocol that proved efficient for the enrichment of PAF and other lysophospholipids such as LPAF and LPC . Tandem MS and bioactivity data obtained in the present study revealed that the T . cruzi PAF-like phospholipid is in fact an LPC , namely sn-1 C18:1 ( Δ9 ) -LPC . However , we do not discard the possibility that very low amounts of a bona fide PAF species might still be synthesized by this parasite . The MSn assignments for the C18:1-LPC we provide here are in agreement with previous reports [54] , [56] , but somewhat different from the results obtained by Smith et al . [41] . Specifically , these authors analyzed a lipid species with a nearly identical fragmentation pattern as what we have identified here as C18:1-LPC , but annotated it as a novel C16:1-alkenyl-PAF . In their study , an ion species at m/z 265 was characterized as being derived from the neutral loss of acetyl and methyl groups from the fragment ion at m/z 339 . In addition , a fragment ion at m/z 247 was proposed to be derived from the loss of water from the ion at m/z 265 . Conversely , we show here that the non-lithiated [R1CO+]+ and [R1CO+ - H2O]+ fragment ions at m/z 265 and 247 , respectively , are in fact derived from a C18:1-acyl chain at the sn-1 position . This is further confirmed by the presence of lithiated ( [R1CO2H+Li]+ ) fragment ion from oleyl chain at m/z 289 . Our assignments of C18:1-LPC are in complete agreement with those reported by Hsu et al . [54] . Smith et al . [41] also proposed the ion species at m/z 223 as being the protonated C16:1-alkenyl chain from the sn-1 position of that putative PAF species . We also observed the same fragment ion in both synthetic and T . cruzi-derived C18:1-LPC , but we assigned it as derived from the fragmentation of the acyl chain . Taking all these facts into consideration , it is likely that Smith et al . [41] have mistakenly identified C18:1-LPC as C16:1-PAF . Here we show that T . cruzi synthesizes at least five species of LPC and that only sn-1 C18:1 ( Δ9 ) -LPC was able to promote rabbit platelet aggregation . These LPCs could be generated by the action of host- and/or parasite-derived phospholipase ( s ) A1 and/or A2 on diacyl-PCs . A PLA1 has already been well characterized in T . cruzi trypomastigote and amastigote forms ( RA , Cvd , and K98 strains ) [79] , [80] and T . brucei [81] . However , since we could only identify sn-1 C18:0- , C18:1- , and C18:2-LPC , we believe that the parasite PLA1 is not playing a major role in the generation of these LPCs , at least under the experimental conditions used in this study . Nevertheless , we could not discard the possibility that the T . cruzi PLA1 might be important for the generation of LPC species using different experimental conditions and/or other parasite strains , as previously described [79] , [80] . Therefore , most likely the LPC species identified in this study were generated by the action of a PLA2 from the host and/or parasite . No PLA2 has so far been purified and fully characterized in T . cruzi , despite the fact that at least four putative PLA2 genes have been annotated in the parasite genome ( TriTrypDB TcCLB . 510743 . 50 , TcCLB . 510659 . 257 , TCSYLVIO_005843 , and Tc_MARK_4470 ) [70] . Although a PLA2-like activity has been previously reported in epimastigotes [82] , no further characterization of this enzymatic activity has been published . Moreover , using more rigorous enzymatic assay conditions , Belauzaran et al . [70] were unable to detect any PLA2 activity in T . cruzi epimastigote supernatants . This raises the possibility that the T . cruzi LPCs identified here could have been generated by the action of a host-derived PLA2 and/or by a parasite-derived enzyme that is expressed in much higher levels in the mammal-dwelling stages . Further studies are needed to address this important point . A direct correlation between increased platelet reactivity and the incidence of acute coronary diseases has been described [83] . Accordingly , there is an increase in platelet aggregation associated with ischemia , myonecrosis , and myocarditis in both acute and chronic stages of Chagas disease [2] , [4] , [5] , [84] . Noteworthy , T . cruzi also synthesizes TXA2 , which modulates several pathophysiological aspects of Chagasic cardiomyopathy [12] . It is not surprising that T . cruzi produces at least two platelet activators , given the importance of platelet aggregation in the progression of Chagas disease . Our bioactivity data with C18:1-LPC are in agreement with other results in the present study , namely the platelet aggregation assays and molecular docking predictions . We had also previously demonstrated that the putative PAF-like phospholipid was able to trigger the differentiation of T . cruzi epimastigotes into metacyclic trypomastigotes in vitro [14] . This effect was abolished by WEB 2086 , a classic competitive PAF antagonist that binds specifically to PAF receptors [85] . The latter result along with labeling T . cruzi epimastigotes with polyclonal antibody raised against mouse PAFR , in immunofluorescence assays , strongly indicated that T . cruzi might have putative PAFR both at the cell surface and intracellularly [14] . In fact , WEB 2086 inhibited all PAF effects upon trypanosomatids described to date [86]–[88] . LPC has messenger functions , binding to a specific receptor and not acting through physicochemical effects on the plasma membrane of the target cell [27] . Several receptors for LPC , such as G2A , GPR4 , and IP , and receptors for TXA2 ( TP ) and PAF ( PAFR ) , have been described [24]–[32] . In the present study , we show that the platelet aggregation promoted by C18:1-LPC was abrogated by WEB 2086 . This result is highly suggestive that T . cruzi-C18:1-LPC may be able to trigger platelet stimulation through a ligand-receptor system . Accordingly , LPC is known to induce intracellular signals [89] and mediate cytokine secretion [90] , both through PAFR . Interestingly , the R . prolixus-derived C16:0-LPC is able to prevent platelet aggregation triggered by PAF [17] , [19] . We may hypothesize that C16:0-LPC acts as a PAF antagonist , binding to its receptor . Indeed , here we show molefcular docking calculations , which are suggestive that C16:0- , C18:0- , and C18:2-LPC could be lodged within the ligand-binding site of PAFR , preventing PAF actions . On the other hand , the predicted binding pose for C18:1-LPC evokes the possibility that this molecule and C16:0-PAF exhibit similar modes of interaction with PAFR . These results can be partially explained by the fact that , depending on the length and degree of unsaturation of the acyl-chain , each LPC species triggers different cellular activities [30] , [34] , [35] . Apparently , each of these LPC species binds to different receptors , probably because of the tridimensional structure of these phospholipids . For instance , TP receptors are involved in the attenuation of vascular relaxation mediated by C18:2- and C20:4-LPC , but not C16:0- or C18:1-LPC [30] . Additionally , similar to C16:0-PAF , C18:1-LPC exhibited the ability to elicit a rapid , receptor-mediated oxidative burst , through generation of reactive oxygen species ( ROS ) in neutrophils . On the other hand , C16:0- and C18:0-LPC did not share the same activity [34] . Consistently , our docking calculations show that both C16:0-PAF and C18:1-LPC structures interact with PAFR model with Asn169 and Thr170 residues via hydrogen bonds . These amino acid residues are localized at the extracellular loop EL2 , which is involved in the ligand recognition and receptor activation in the superfamily of the GPCRs [91] . The number and the chain length of the fatty acids of lipoproteins are also substantially important for the induction of signaling through Toll-like receptor 2 ( TLR2 ) [92] . A cross-talk between TLR and GPCR has been described , which is essential for cellular signaling in the absence of TLR natural ligands . This cross-talk is basically a molecular organizational GPCR signaling platform that promotes the transactivation of TLR , through potentiation of Neuraminidase 1 and matrix metalloproteinase-9 at the cell surface [93] . Intriguingly , LPC derived from the human pathogen Schistosoma mansoni activates TLR-2-dependent signaling involved in eosinophil activation and recruitment , probably through cross-talk to a GPCR [94] . More recently , we have shown that various species of LPC containing different acyl chain lengths and degrees of unsaturation may induce proinflammatory response mediated by TLR4- or TLR2/TLR1-dependent signaling pathway in HEK293A cells [95] . Interestingly , a mixture of LPC species ( mainly C16:0- and C18:0-LPC ) could counteract TLR4-mediated signaling pathway triggered by E . coli O111:B4 lipopolysaccharide ( LPS ) , suggesting therefore a dual role of this lysophospholipid in immunoregulation . Over time , LPC was found not to be exclusively of mammalian origin and was described in several other organisms [96] . Trypanosomatids tend to be rich in PC and LPC [36]–[39] , [97] . However , in most of these cases , the presence of LPC has only been inferred indirectly and there still remains a general lack of information in the literature regarding the chemical structure and function of the LPC species in most organisms other than mammals . The results described in the present study show that T . cruzi is able to generate different LPC species and suggest that the levels of individual LPC species are tightly regulated in the course of this parasite cell cycle . In fact , the LPC levels are notably high in the mammal-dwelling infectious trypomastigote and intracellular amastigote stages . Taken together with the fact that there was no exogenous source ( e . g . , fetal bovine serum ) of lipids in the TAU culture medium used for the differentiation of epimastigotes into metacyclic trypomastigotes , our data indicate that LPC species might be generated endogenously by this organism . The saliva of R . prolixus is a source of LPC , which may act as an enhancing factor for Chagas disease [17] , [18] . LPC , TXA2 , and PAF are lipid mediators that share a common chemical structure , biosynthetic pathways [98] and the ability to activate platelets [99]–[101] . Previous results from our group have shown that PAF-treated T . cruzi is far more infective towards both mouse macrophages [14] and the insect vector R . prolixus [102] . TXA2 synthesized by T . cruzi controls the proliferation of the parasite and resulting inflammatory response to infection in the mouse [12] . One may then speculate that host-vector-parasite co-evolutionary relationships may be involved in the maintenance or change in the enzymatic pathways for the synthesis and degradation of these molecules , which ultimately may dictate the success of T . cruzi infection . In conclusion , here we demonstrate that T . cruzi synthesizes at least five LPC species , C16:0- , C18:0 , C18:1- , C18:2- , and C22:6-LPC . The most abundant species are C18:2- and C18:1-LPC , being the latter the only one able to aggregate rabbit platelets , probably through a PAFR . This result was supported by molecular docking study of the interactions between LPC species and a PAFR model . These analyses showed that C18:1-LPC was able to interact with the PAFR model in a fashion similar to PAF . More studies on LPC metabolism in T . cruzi and relationship of the parasite with the mammalian and invertebrate hosts could , therefore , lead to the discovery of putative targets for novel therapies and control for Chagas disease .
|
Chagas disease , caused by the parasite Trypanosoma cruzi , was exclusively confined to Latin America but it has recently spread to other regions of the world . Chagas disease affects 8–10 million people and kills thousands of them every year . Lysophosphatidylcholine ( LPC ) is a major bioactive phospholipid of human plasma low-density lipoproteins ( LDL ) . Platelet-activating factor ( PAF ) is a phospholipid similar to LPC and a potent intercellular mediator . Both PAF and LPC have been reported to act on mammalian cells through PAF receptor ( PAFR ) . Previous data from our group suggested that T . cruzi produces a phospholipid with PAF activity . Here , we describe the structural and functional analysis of different species of LPC from T . cruzi , including a LPC with a fatty acid chain of 18 carbon atoms and one double bond ( C18:1-LPC ) . We also show that C18:1-LPC is able to induce rabbit platelet aggregation , which is abrogated by a PAFR antagonist . In addition , a three-dimensional model of human PAFR was constructed . Contrary to other T . cruzi LPC molecules , C18:1-LPC is predicted to interact with the PAFR model in a fashion similar to PAF . Further studies are needed to validate the biosynthesis of T . cruzi C18:1-LPC as a potential drug target in Chagas disease .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"biochemistry",
"lipids",
"lipid",
"mediators",
"biology",
"and",
"life",
"sciences",
"microbiology",
"parasitology"
] |
2014
|
Structural and Functional Analysis of a Platelet-Activating Lysophosphatidylcholine of Trypanosoma cruzi
|
Yeast pseudohyphal filamentation is a stress-responsive growth transition relevant to processes required for virulence in pathogenic fungi . Pseudohyphal growth is controlled through a regulatory network encompassing conserved MAPK ( Ste20p , Ste11p , Ste7p , Kss1p , and Fus3p ) , protein kinase A ( Tpk2p ) , Elm1p , and Snf1p kinase pathways; however , the scope of these pathways is not fully understood . Here , we implemented quantitative phosphoproteomics to identify each of these signaling networks , generating a kinase-dead mutant in filamentous S . cerevisiae and surveying for differential phosphorylation . By this approach , we identified 439 phosphoproteins dependent upon pseudohyphal growth kinases . We report novel phosphorylation sites in 543 peptides , including phosphorylated residues in Ras2p and Flo8p required for wild-type filamentous growth . Phosphoproteins in these kinase signaling networks were enriched for ribonucleoprotein ( RNP ) granule components , and we observe co-localization of Kss1p , Fus3p , Ste20p , and Tpk2p with the RNP component Igo1p . These kinases localize in puncta with GFP-visualized mRNA , and KSS1 is required for wild-type levels of mRNA localization in RNPs . Kss1p pathway activity is reduced in lsm1Δ/Δ and pat1Δ/Δ strains , and these genes encoding P-body proteins are epistatic to STE7 . The P-body protein Dhh1p is also required for hyphal development in Candida albicans . Collectively , this study presents a wealth of data identifying the yeast phosphoproteome in pseudohyphal growth and regulatory interrelationships between pseudohyphal growth kinases and RNPs .
The pseudohyphal growth response is a complex morphogenetic program in which fungal cells transition from a yeast-like growth form to a filamentous state , with cells remaining physically connected after cytokinesis in elongated structures [1–3] . This growth transition is evident in several strains of S . cerevisiae ( e . g . , Σ1278b and SK1 ) [4 , 5] and is triggered by numerous conditions , including nitrogen limitation , glucose limitation , the presence of starch as a sole carbon source , and elevated levels of fusel alcohols [1 , 6–9] . Since yeast pseudohyphal growth is principally induced in response to nutrient stress , it is widely presumed to be a nutritional foraging mechanism [10] . Pseudohyphal growth has been studied intensely in S . cerevisiae as an informative model of related processes of filamentous growth evident in many fungi . In particular , the pseudohyphal growth transition in S . cerevisiae is closely related to filamentous growth transitions enabling the formation of pseudohyphae and true hyphae with parallel-sided cell walls in the opportunistic human fungal pathogen Candida albicans [11–13] . Further , the ability to form hyphae and to transition between these growth forms is required for virulence in C . albicans [14–16] . The molecular basis of yeast pseudohyphal growth is extensive . Pseudohyphal formation in S . cerevisiae is enabled by changes in cell polarity , cytoskeletal organization , and cell adhesion controlled through a regulatory network encompassing a core set of strongly conserved signaling modules [17–20] . Yeast cells contain several mitogen-activated protein kinase ( MAPK ) pathways , and elegant studies in the mid-1990s identified the cascade of Ste11p , Ste7p , and Kss1p as a pseudohyphal growth activator [21–23] . Within this pseudohyphal growth MAPK pathway , the upstream p21-activated kinase Ste20p phosphorylates and activates Ste11p , and this phosphorylation signal is propagated through Kss1p to the heterodimeric transcription factor Ste12p/Tec1p [24 , 25] . Ste11p and Ste7p are also components of a pheromone-responsive MAPK cascade containing the MAPK Fus3p [26 , 27] . Fus3p negatively regulates pseudohyphal growth by phosphorylating Tec1p Thr273 , targeting Tec1p for degradation [28] . In addition to these MAPK pathways , cAMP-dependent protein kinase A ( PKA ) is a key regulator of pseudohyphal development . In S . cerevisiae , PKA consists of the Bcy1p regulatory subunit and one of three catalytic subunits , Tpk1p , Tpk2p , and Tpk3p [29 , 30] . Tpk2p phosphorylates the filamentous growth transcriptional activator Flo8p , and deletion of TPK2 reduces pseudohyphal growth [31 , 32] . The AMP-activated kinase Snf1p is a well-studied transcription factor required for derepressed expression of glucose-repressible genes [33] . Snf1p represses the pseudohyphal growth negative regulators Nrg1p and Nrg2p , resulting in transcriptional activation of FLO11 , among other targets [34] . FLO11 encodes a GPI-anchored cell surface flocculin required for pseudohyphal growth , acting as a downstream effector of the Kss1p MAPK pathway through Ste12p/Tec1p , the PKA pathway through Flo8p , and Snf1p as described [35 , 36] . Snf1p Thr210 is phosphorylated by Elm1p , which regulates cellular morphogenesis and cytokinesis [37] . The core components of these signaling pathways are well established , but the set of targets of each signaling module are not as clearly defined with respect to the gene network contributing to pseudohyphal growth . Systematic analysis of loss-of-function mutants revealed that approximately 700 genes are required for wild-type pseudohyphal growth [38 , 39] , and a partially overlapping set of 551 genes promotes invasive growth upon galactose-induced overexpression [40] . In particular , the regulation of stress-responsive processes during pseudohyphal growth is a point of ongoing study , indicating counterbalanced control of autophagy through Tor/PKA [41] and an extensive glucose-regulated signaling network encompassing Snf1p and related pathways [34 , 42] . Pseudohyphal growth gene networks have been analyzed globally for regulatory control at the level of transcription [43 , 44] , but kinase signaling networks regulating filamentous growth have been constructed predominantly from individual studies of a given kinase and target . Although kinase signaling in yeast has been analyzed effectively through mass spectrometry-based phosphoproteomics [45–48] , these methods had not been applied to define kinase networks controlling pseudohyphal growth in a filamentous strain of S . cerevisiae . Here , we implemented quantitative phosphoproteomics to identify signaling networks for a set of kinases that regulate filamentation , with the results revealing a wealth of previously unknown phosphorylation sites , phosphorylated residues in Ras2p and Flo8p required for pseudohyphal growth , and MAPK regulation of ribonucleoprotein complexes via the Kss1p cascade .
To dissect kinase signaling networks regulating yeast pseudohyphal growth , we adopted a straightforward approach , generating a loss-of-function mutation in relevant pseudohyphal growth kinases and surveying the resulting changes in phosphorylation . For this study , we constructed catalytically impaired kinase-dead alleles in the filamentous Σ1278b strain of S . cerevisiae for each of the following kinases: Ste20p , Ste11p , Ste7p , Kss1p , Fus3p , Tpk2p , Elm1p , and Snf1p . The signaling context of each protein is indicated in S1A Fig . Mutant kinase alleles were generated by deletion of each kinase gene and introduction of a low-copy centromeric plasmid bearing the native gene promoter and mutated coding sequence encoding a Lys-to-Arg substitution at the conserved residue in the catalytic loop of each respective kinase domain . Resulting filamentous growth phenotypes are presented in S1 Table; images of these kinase-dead mutants as well as background deletion strains and isogenic strains carrying wild-type kinase genes are shown in S1B–S1D Fig . Differential phosphorylation in the kinase-dead mutants relative to wild type was assessed by quantitative phosphoproteomics using stable isotopic labeling of amino acids in cell culture ( SILAC ) [49] . To implement this SILAC-based approach , the wild type and kinase mutant strains in Σ1278b were made auxotrophic for lysine and arginine by deletion of LYS1 and ARG4 . The resulting strains were cultured in triplicate under conditions inducing pseudohyphal growth in medium containing either isotopically labeled or unlabeled lysine and arginine . Protein extracts from the cultures were enriched for phosphopeptides , and the enriched fractions were analyzed by liquid chromatography coupled with tandem mass spectrometry to determine the identity and relative abundance of the phosphopeptides . This experimental design and workflow is summarized in Fig 1A . By SILAC-based phosphoproteomics , we identified 11 , 337 peptide-to-spectrum matches and filtered the peptides by criteria indicated in Experimental Procedures . The resulting data indicate 3 , 699 unique phosphopeptides , corresponding to 1 , 111 proteins . In total , we observed 711 peptides that exhibited a change in phosphorylation ( SILAC ratio ≥1 . 5 or ≤0 . 5 with statistical significance ≤0 . 05 ) in the respective kinase mutants relative to wild type ( Fig 1B ) . This differentially phosphorylated peptide set corresponds to 439 phosphoproteins . A full listing of these data is provided in S2 Table . In addition to Ser/Thr phosphorylation , we also identify 137 peptides with a phosphorylated tyrosine residue , corresponding to 110 yeast proteins . By our experimental design , the proteins identified in this study encompass both direct and indirect targets of the respective pseudohyphal growth kinases . These phosphoproteins encompass eighteen functionally uncharacterized proteins and 73 proteins whose corresponding genes yield pseudohyphal growth phenotypes upon deletion ( Fig 1B ) [38 , 39] . These mass spectrometry studies provide a substantial catalog of previously unreported phosphorylation sites across the yeast proteome . Due to the lack of a centralized repository of previously identified phosphopeptides , novel sites were identified through a multi-step process: we first generated a compendium of known phosphorylation sites culled from phosphopeptide databases ( Materials and Methods ) , and we subsequently mapped peptides and phosphorylation sites from each of these databases onto the yeast proteome along with phosphosites identified in our data . Allowing for inherent uncertainties in both our data and reported phosphorylation sites from the community databases , we identified sites that were well distinct from those previously reported . A listing of potentially novel phosphorylation sites can be accessed from S2 Table . Interestingly , as indicated in Fig 2 , we identified previously unreported phosphorylation sites in the GTP-binding protein Ras2p ( Y165 , T166 ) and the pseudohyphal growth transcription factor Flo8p ( S587 , S589 , S590 ) , both proteins being required for filamentous growth [50 , 51] . Mutation of these sites to encode non-phosphorylatable residues results in decreased invasive growth ( S2 Fig ) . A flo8 mutant encoding alanine at residues 587 , 589 , and 590 ( flo8-S3A ) results in decreased production of lacZ driven from a segment of the FLO11 promoter containing Flo8p-binding sites ( Pflo11-6/7 ) . Additionally , the flo8-S3A allele yields decreased activity of a lacZ reporter driven by a filamentation-responsive element ( FRE ) recognized by the Kss1p-regulated Ste12p/Tec1p transcription factor [22] ( Fig 2B ) . The ras2-Y165F/T166A and flo8-S3A alleles encode proteins that can be visualized as GFP chimeras exhibiting wild-type localization to the plasma membrane and nucleus , respectively ( S2 Fig ) . To identify cellular processes involved in the pseudohyphal growth transition , we mined the collective set of proteins differentially phosphorylated in the kinase-dead mutants for statistically significant enrichment of associated Gene Ontology ( GO ) terms using the Biological Process , Molecular Function , and Cellular Component vocabularies . In addition to the expected identification of terms associated with pseudohyphal growth and polarized growth , this analysis indicated enrichment for proteins involved in translational regulation ( Fig 3A ) . Using the DAVID bioinformatics suite , we identified a cluster of genes annotated with related GO terms involving the regulation of translation ( GO ID:0006417 ) , the regulation of cellular protein metabolic process ( GO ID:0032268 ) , and the posttranscriptional regulation of gene expression ( GO ID:0010608 ) . The gene set contains protein components of mRNA-protein granules , and gene sets annotated with the GO terms RNP granule ( GO ID:00035770 ) , cytoplasmic mRNA processing body ( GO ID:0000932 ) , and cytoplasmic stress granule ( GO ID:0010494 ) are statistically enriched in the set of differentially phosphorylated proteins observed for many of the kinase-dead mutants ( Fig 3B ) . It is noteworthy that the subset of identified proteins exhibiting increased phosphorylation in the kinase-dead strains , presumably encompassing indirect kinase targets , is not enriched for GO terms associated with RNP granules . A listing of GO terms enriched in this hyper-phosphorylated protein subset is presented in S3 Table . Cytoplasmic stress granules and mRNA processing bodies ( P-bodies ) are two classes of RNPs observed in yeast , with compositional and presumed functional similarities to large families of RNP particles observed throughout eukaryotes . As reviewed in Buchan and Parker [52] , the stress-induced RNPs contain non-translating mRNAs and function in an mRNP cycle , wherein mRNA may traffic between mRNPs that exhibit a dynamic protein makeup . Classically , P-bodies are thought to be aggregates of mRNA with proteins involved in translational repression , deadenylation , decapping , and 5’-to-3’ exonucleolytic mRNA decay [53 , 54] , while the protein composition of yeast stress granules encompasses translation initiation factors suggestive of associated RNAs stalled in translation initiation [55 , 56] . Our phosphoproteomic analysis identifies pseudohyphal growth ( PHG ) kinase-dependent phosphorylation of proteins localized in P-bodies ( Dcp2p , Ded1p , Dhh1p , Edc3p , Pat1p , Sbp1p , and Xrn1p ) and stress granules ( Eap1p , Hrp1p , Pbp1p , and Ygr250cp ) , as well as proteins identified in both ( Igo1p , Ngr1p , and Tif4632p ) [57] . A listing of PHG kinase-dependent phosphorylation sites in these RNP granule proteins is indicated in Table 1 . As a further step towards identifying a regulatory link between pseudohyphal growth kinase signaling and RNP biology , we constructed a network connectivity map integrating physical interactions between: 1 ) RNP components and 2 ) proteins in signaling pathways/cell processes required for filamentation . Using annotations from the Kyoto Encyclopedia of Genes and Genomes ( KEGG ) , we identified a fairly dense map between proteins localized to RNPs and the indicated KEGG pathways related to filamentous growth ( Fig 3C ) . We did not observe maps of similar density between RNP components and proteins in other filamentous growth-related KEGG pathways . The proteins used to generate this network and the database sources of the interactions are listed in S4 Table . Since the phosphorylation of numerous RNP components is dependent upon pseudohyphal growth kinases , we examined the set of eight kinases selected here for co-localization with RNP particles . For this analysis , we generated carboxy-terminal GFP-fusions to each kinase and carboxy-terminal mCherry fusions to several RNP components ( Dcp2p , Edc3p , Igo1p , and Pat1p ) as chromosomal alleles in a strain of the filamentous Σ1278b genetic background . As indicated in Fig 4A , Fus3p , Kss1p , Ste20p ( in the MAPK pathway ) and Tpk2p ( PKA ) co-localized as GFP chimeras with Igo1p-mCherry . Igo1p is a RNP component of unknown function that has been identified as a Rim15p target required for proper initiation of G0 [58] . Rim15p , acting downstream of PKA , phosphorylates Igo1p at S64; this phosphorylation site was also observed in the mass spectrometry studies reported here . Igo1p binds the P-body protein Dhh1p and the stress-granule protein Pbp1p as determined previously by co-immunoprecipitation [58] . We observe Igo1p-mCherry puncta and co-localization with Fus3p , Kss1p , Ste20p , and Tpk2p post-diauxic shift , upon 3 days growth in standard media ( Fig 4A ) . We further considered co-localization of pseudohyphal growth kinases with RNPs using the RNA visualization strategy of Brodsky and Silver [59] . By this method , RNA can be visualized in puncta by fluorescence microscopy of a U1A-GFP fusion bound to multiple U1A-binding sites introduced into the 3’-untranslated region of a target mRNA . For this study , we used PGK1 mRNA modified with sixteen U1A-binding sites as a marker of bulk and stable mRNAs ( Fig 4B ) . As described in Sheth and Parker [60] , this fluorescence-tagging strategy has been used to visualize RNPs as puncta , and we have co-localized Pbp1-GFP with U1A-mCherry-bound RNA puncta by this approach . Each pseudohyphal growth kinase tested in this study was analyzed for co-localization with PGK1 RNA puncta , and , consistent with the results presented above , we observed substantial RNA puncta co-localization with mCherry fusions to Fus3p , Kss1p , Ste20p , and Tpk2p ( Fig 4B ) . In contrast , mCherry chimeras with Elm1p , Snf1p , Ste7p , and Ste11p did not co-localize with the engineered PGK1 RNA . Kinase puncta were evident under a short period of glucose limitation , with RNA co-localization persistent through at least two days . Transferring cells from media limited in glucose to media with normal levels of glucose resulted in a loss of observed puncta , indicating that punctate formation was responsive to glucose limitation . The RNP protein Igo1p is PKA-regulated , and we find that wild-type localization of the PKA catalytic subunit Tpk2p requires IGO1 and its paralog IGO2 . The IGO1 and IGO2 genes arose from a whole genome duplication event , with 58% identity between these proteins [58] , suggesting that these genes are functionally redundant . Upon deleting both genes , complex colony morphology is exaggerated , consistent with perturbed PKA signaling [61] ( Fig 5A ) . IGO1 and IGO2 are required for signaling through the Kss1p MAPK pathway and PKA pathway as assessed using lacZ reporters containing FRE sites responsive to Kss1p-regulated Ste12p/Tec1p and a segment of the FLO11 promoter bound by PKA-regulated Flo8p , respectively ( Fig 5B ) . Further , the punctate localization of a Tpk2p-GFP chimera is disrupted in a strain deleted for IGO1 and IGO2 ( Fig 5C ) . The subcellular distribution of Kss1p , Fus3p , and Ste20p is unaffected in an igo1Δ/igo2Δ mutant in the filamentous Σ1278b background ( S3C Fig ) . Numerous genes encoding RNP components are required for wild-type pseudohyphal growth . We identified strongly diminished pseudohyphal growth in strains of the filamentous Σ1278b background containing homozygous deletions of the mRNA decay/translational repressor genes CCR4 , DHH1 , LSM1 , PAT1 , PBP1 , and SBP1 ( Fig 5D ) . To assess activity of the MAPK pathway in these mutants , we introduced the plasmid-based FRE-lacZ Kss1p reporter and assayed for β-galactosidase activity under conditions of nitrogen limitation . As indicated in Fig 5D , lsm1Δ/Δ and pat1Δ/Δ mutants were strongly decreased in MAPK pathway activity relative to wild type , approaching levels observed in a homozygous diploid strain deleted for STE12 . LSM1 and PAT1 both encode P-body proteins that function in mRNA decapping [60]; Lsm1p functions with a group of six other Lsm family proteins in a complex that associates with the Pat1p decapping enzyme , and , collectively , the proteins mediate mRNA decay through decapping [62] . Epistasis studies indicate that overactive alleles of STE11 and STE7 [63] are suppressed by deletion of LSM1 and PAT1 , both morphologically and by FRE-lacZ reporter assay ( Fig 5E ) . A hyperactive KSS1 allele did not yield an exaggerated pseudohyphal growth phenotype . These results indicate that the core mRNA decapping proteins Lsm1p and Pat1p are required for wild-type pseudohyphal growth MAPK signaling , and that the genes act at or below the level of the MAPKK STE7 . In addition to the requirement for RNP components in wild-type Kss1p signaling , we find that KSS1 is required to achieve wild-type levels of mRNA puncta . A strain of the filamentous Σ1278b background deleted for KSS1 yields a significantly decreased amount of U1A-GFP-tagged PGK1 mRNA localized in puncta under conditions of glucose limitation ( Fig 5F ) . This phenotype was consistent in the kss1Δ strain from a brief 15-minute incubation under conditions of glucose limitation up to a period of at least eight hours . The percentage of cells exhibiting mRNA puncta decreased from 65% in wild-type cells to 15% in kss1Δ mutants , and , correspondingly , the maximum number of puncta observed in these cells decreased from maximally 16 in wild-type to no greater than four in the kss1Δ strain . Visible puncta were lost upon the introduction of media with normal levels of glucose . Profiling studies indicate that bulk RNA association with polyribosomes is not substantially altered in a strain of the Σ1278b background deleted for KSS1 ( S4 Fig ) , although the translational processing of specific transcripts may still be perturbed in a kss1Δ strain . To consider the likelihood of a conserved functional interrelationship between RNP components and filamentous growth , we assessed the contributions of the P-body protein Dhh1p towards hyphal development in the pathogenic fungus Candida albicans . We selected DHH1 for study because it is a core component of P-bodies , and its localization has been confirmed in C . albicans . Further , in S . cerevisiae , a loss-of-function mutation in DHH1 results in decreased pseudohyphal growth ( Fig 5D ) , although its hyphal growth phenotype in C . albicans has not been clearly identified . The C . albicans ortholog of DHH1 is presumed to function similarly to S . cerevisiae DHH1 , and the genes exhibit 91% sequence similarity at the encoded amino acid level . For this analysis , we generated a heterozygous deletion of DHH1 in C . albicans by standard gene replacement and assayed for altered colony and cell morphology under conditions inducing hyphal development . As indicated in Fig 6A , the dhh1Δ/DHH1 strain exhibits reduced surface wrinkling and peripheral hyphae relative to an isogenic wild-type strain . Deletion of DHH1 resulted in cells that were less elongated than wild type , and , correspondingly , hyphae were decreased in number under these conditions ( Fig 6B ) .
The phosphoproteomic analysis presented here is the first such study of kinase signaling networks in pseudohyphal growth , effectively complementing previous phosphoproteomic analyses of yeast kinases in a non-filamentous strain under standard growth conditions . Compared to data in the major phosphorylation databases ( Experimental Procedures ) , this analysis identifies a large set of previously unreported phosphorylation sites . These sites encompass residues that are phosphorylated strictly in the Σ1278b proteome in response to the conditions employed here as well as phosphorylation sites that were not sampled in previous analyses . This suggests that the quantitative phosphoproteomic data collected here and elsewhere are not saturating . Consequently , there is benefit in continued analysis of the yeast phosphoproteome towards understanding kinase networks more fully . Our mass spectrometry data identify 73 proteins that: 1 ) were differentially phosphorylated in a kinase-dead strain , and 2 ) result in a pseudohyphal growth phenotype upon deletion . Ryan et al . [39] reported 497 genes that result in diploid pseudohyphal growth defects in S . cerevisiae under conditions of low nitrogen; thus , the kinase signaling network identified here encompasses a significant fraction of this gene set . The network contains both direct and indirect kinase targets . In a landmark study , Ptacek et al . [64] used a proteome microarray to identify in vitro substrates of most yeast kinases . We integrated the results from this microarray study for the kinases tested here with our own data , identifying in vitro substrates that also exhibited differential phosphorylation by mass spectrometry . The results are indicated in S5 Table . Further interpretation of the kinase-dependent phosphorylation data yields two observations . First , many of the identified phosphorylation sites were dependent on the presence of more than one pseudohyphal growth kinase . This observation holds true for kinases that are thought to function in distinct pathways . Of 311 unique proteins identified as undergoing differential phosphorylation in a strain carrying a kinase-dead mutation in the MAPK Kss1p , 188 of these proteins were also differentially phosphorylated in a strain carrying a kinase-dead allele of ELM1 . We report 814 unique proteins differentially phosphorylated in tpk2-K99R , and 249 of these proteins were also differentially phosphorylated in a strain with a kinase-dead allele of KSS1 . Collectively , this suggests that the respective kinases and pathways regulate partially overlapping signaling networks . Overlapping signaling networks further suggest a degree of functional redundancy , although mutant phenotypes are evident upon deletion of each individual kinase analyzed here . Second , the identification of new and functionally important phosphorylated residues in pseudohyphal growth proteins such as Flo8p underscores the utility in extending quantitative phosphoproteomic studies to the analysis of non-standard growth conditions and strains , as FLO8 is an incompletely translated pseudogene in standard S288C laboratory strains of S . cerevisiae . The functional interrelationship identified in this study between RNP components and pseudohyphal growth is supported by several lines of evidence obtained in the non-filamentous S288C genetic background . As reported in Yoon et al . [65] , Ste20p phosphorylates Dcp2p at Ser 137 , and this phosphorylation is required for Dcp2p localization in P-bodies . Shah et al . [66] found that overexpression of PKA isoforms inhibits P-body formation and that the Tpk1p and Tpk2p subunits of PKA are capable of phosphorylating the P-body protein Pat1p in vitro . Also in an S288C background under conditions of glucose depletion , Xrn1p undergoes Snf1p-dependent phosphorylation along with a subset of additional mRNA processing proteins [48] . In S288C , Fus3p co-localizes with P-bodies and stress granules as yeast cells enter stationary phase [67] . It is noteworthy that Kss1p is non-functional in the S288C strain [26] and presumably would not have been identified as a regulator of RNPs in previous studies undertaken in that genetic background . In the filamentous Σ1278b strain , the kinases Kss1p , Fus3p , Tpk2p , and Ste20p co-localize with PGK1 mRNA foci as well as with the P-body and stress-granule protein Igo1p . As presented in Buchan et al . [68] , mRNAs are thought to traffic between P-bodies , stress granules , and other RNPs , with the composition of each particle being dynamic in response to the specific cell stress and duration of the stress . Since we do not at present understand the full composition of these particles in a filamentous strain under the observed growth conditions , we use the term RNP here to indicate co-localization with RNA/protein foci and specifically identify co-localization with Igo1p . It is interesting that the upstream PAK Ste20p and the MAPK Kss1p co-localize with Igo1p and GFP-tagged mRNA , but neither the MAPKKK Ste11p nor the MAPKK Ste7p share this localization pattern . Three points are relevant in considering the localization of these pathway components . First , prior literature supports the observation that MAPK pathway components may not be uniformly localized . For example , in the mating pathway , Ste7p and Fus3p have been identified at the bud tip , although Ste11p has not been similarly localized . Fus3p exists in several complexes affecting its localization [69] . Kss1p has been reported previously to localize to the nucleus [70] , although neither Ste11p nor Ste7p are found predominantly in the nucleus . Second , this work and other studies [64 , 71] indicate that the respective MAPK pathway components do not exhibit strictly overlapping targets . The fact that these kinases recognize nonoverlapping sets of targets suggests that differential localization of subpopulations of the kinases is possible . Third , Kss1p and Fus3p as well as the pseudohyphal growth upstream PAK Ste20p are not exclusively identified in mRNPs , and the likelihood exists that a subset of these respective protein populations may indeed be colocalized . RNPs are induced in response to numerous cell stresses , including glucose limitation , hyperosomotic stress , and high cell density [72–74]; however , we find that nitrogen stress , a classic inducer of pseudohyphal growth , is not a strong inducer of RNPs in the absence of additional cell stresses . The mechanisms of these inductions are unclear; consequently , it is also unclear as to why nitrogen limitation alone is insufficient to strongly induce this response . In this analysis , 1-butanol , rather than low-nitrogen media , was used as an inducer of filamentation because of its ability to yield a strong filamentous response in liquid cultures . Further , the cells necessitated growth to 10 doublings under conditions inducing pseudohyphal growth to achieve efficient labeling , which resulted in a mild degree of glucose stress . The presence of short-chain alcohols coupled with glucose limitation provides strong induction of filamentation in liquid . Notably , we observe RNP foci under these conditions for a growth period of at least 30 hours , representing a time point matching the endpoint of growth for quantitative phosphoproteomic analysis . It should be noted that stress granules form post entry into stationary phase [66] , and consequently , this analysis may be less effective in identifying post-translational modifications affecting stress granule components . In sum , however , the conditions present at the point of mass spectrometric analysis allow for the presence of RNPs , consistent with the identification of differentially phosphorylated RNP components in this study . Interestingly , the requirement for DHH1 , encoding a helicase involved in mRNA decapping , is conserved between S . cerevisiae and the related pathogenic fungus C . albicans . Signaling pathways in S . cerevisiae serve as effective models of related pathways in C . albicans [75] , raising the possibility that RNPs and hyphal development are functionally linked in C . albicans as well . In support of this notion , P-bodies have been observed to form during hyphal development in C . albicans , and Dhh1p does co-localize to P-bodies in Candida [76] . Further , a strain of C . albicans deleted for EDC3 exhibits a defect in filamentation [76] . The results here are consistent with regulatory feedback between RNP components and hyphal development in this opportunistic human pathogen . In total , the data support an interrelationship between pseudohyphal growth kinase signaling and RNP biology . Deletion analyses and epistasis studies indicate that RNA processing proteins are required for wild-type Kss1p MAPK signaling , likely to regulate the translational state of particular transcripts important for pseudohyphal growth . In turn , the identified pseudohyphal growth kinases localize to RNPs , and the Kss1p pathway is required for wild-type RNP numbers in the filamentous S . cerevisiae strain . Thus , Kss1p MAPK signaling and RNP signaling feed back reciprocally . Similarly , PKA regulates Igo1/2p function through Rim15p , and Igo1/2p is in turn required for wild-type PKA localization . The data here collectively provide an important step towards identifying the mechanisms through which this reciprocal signaling is mediated .
Strains used in this study are listed in S6 Table . S . cerevisiae strains were derived from the filamentous Σ1278b genetic background . Haploid strains were derived from Y825 and HLY337 [1 , 77] . Standard protocols and techniques were used for the propagation of budding yeast as described [78] . DNA was introduced by methods of yeast transformation incorporating lithium acetate treatment and heat shock [79] . Plasmids used in this study are listed in S7 Table . S . cerevisiae strains were cultured on YPD ( 1% yeast extract , 2% peptone , 2% glucose ) or Synthetic Complete ( SC ) ( 0 . 67% yeast nitrogen base ( YNB ) without amino acids , 2% glucose , and 0 . 2% of the appropriate amino acid drop-out mix ) . Nitrogen deprivation and filamentous phenotypes were assayed using Synthetic Low Ammonium Dextrose ( SLAD ) medium ( 0 . 17% YNB without amino acids , 2% glucose , 50 μM ammonium sulfate and supplemented with appropriate amino acids ) or by supplementing growth medium with 1% 1-butanol [9] . Glucose limitation was achieved using media lacking glucose as a carbon source according to standard protocols . Gene deletions and tags for chromosomal integration were generated through one-step PCR- mediated transformation and subsequent PCR-based verification [80 , 81] . N-terminal and C-terminal GFP tagging was performed using plasmid-based modules from Longtine et al . [82] . Carboxy-terminal mCherry tagging was performed by PCR-based amplification of the mCherry-kanMX or HygR cassette of pBS34 or pBS35 ( Yeast Resource Center , Univ . of Washington ) . Integrated point mutations in RAS2 ( Y165F , T166A ) and FLO8 ( S587A , S589A , and S590A ) were generated by the URA “flip-out” method . In brief , the URA3 cassette of pRS406 was amplified by PCR and used to disrupt local sequence at the intended site of mutation . A second transformation was then performed to replace the URA3 marker with DNA sequence encoding the desired mutated allele [5] . ARG4 and LYS1 were deleted in the haploid Y825 filamentous strain . Subsequently , protein kinase genes were individually deleted in the filamentous arg4Δ lys1Δ auxotrophic Y825 background . These arginine/lysine auxotrophic kinase null strains were then transformed with LEU2-containing yeast shuttle vectors carrying the respective kinase-dead alleles . SILAC was used to differentially label proteins synthesized by kinase-dead allele strains and wild-type ( arg4Δ lys1Δ Y825 ) . SILAC-based mass spectrometry experiments were multiplexed; each mass spectrometry experiment was conducted in triplex , with the light ( natural ) versions of L-arginine and L- lysine used to label the wild-type strain and medium ( Lys-4/Arg-6 ) or heavy ( Lys-8/Arg-10 ) L-arginine and L-lysine labeling two kinase-dead allele strains . In each triplex experiment , three strains were cultured in parallel during filamentous growth-inducing conditions . Wild-type and two kinase-dead allele strains were cultured overnight in synthetic complete media containing arginine or lysine residues with light , medium , or heavy isotopes overnight at 30°C , to obtain actively growing log-phase cultures . Each culture was then diluted to a low starting optical density ( OD600 of approximately 0 . 1 ) in SILAC media . To induce filamentous growth in these haploid strains , 1% ( vol/vol ) 1-butanol was added to each culture . These diluted cultures were subsequently incubated at 30°C for approximately 10 doublings ( approximately 26 hours ) . This prolonged labeling and culturing step was found to be necessary to ensure effective metabolic labeling of proteins , as well as to obtain a minimum abundance of labeled protein from each strain . Protein extractions and mass spectrometry were performed as described previously [42 , 83] . We processed mass spectrometry data using maxQuant [84] and collated the list of phospho ( STY ) peptides . The data were filtered at 5% FDR; additionally , we excluded peptides exhibiting low Mascot scores ( <3 ) , high charge states ( ≥5 ) , and long peptide lengths ( >40 ) . A normalized heavy:light or medium:light ratio with a significance score ( Sig A ) ≤ 0 . 05 was considered statistically significant . Predicted phosphorylation sites were screened for known kinase motifs and the results from this analysis are included in the “Motifs” column in S2 Table . To identify potentially novel phosphorylation sites , Exonerate was used to align peptides extracted from major phosphorylation databases ( GPM DB , PhosphoPep , PHOSIDA , and Phospho . ELM ) onto yeast protein sequence data from the Saccharomyces Genome Database . Phosphosites were marked on the protein sequences , yielding a compendium of phosphorylation sites . Inherent ambiguities in the localization of phosphorylation sites were annotated as such in the compendium . Phosphorylation sites identified in our data were subsequently mapped onto the annotated phosphoproteome . Networks were constructed using background sets of protein interactions encompassing kinase-dependent differentially phosphorylated proteins in the mass spectrometry data generated here as well as interactions identified in the iRefIndex database . KEGG signaling pathways relevant for pseudohyphal growth were downloaded and parsed using in-house scripts . The resulting network was expanded by including previously identified core components of stress granules and P-bodies [57 , 85] . The network was visualized using Cytoscape . The interactions used to construct this network and the database source of each interaction are provided in S4 Table . RNA localization in RNP foci was visualized as described [57] . Live S . cerevisiae cells with these plasmids and/or fluorescent protein fusions to known mRNP components were imaged using an upright Nikon Eclipse 80i microscope with CoolSnap ES2 CCD ( Photometrics ) . Images were acquired using the MetaMorph software package ( Molecular Devices ) . Candida albicans strains used in this study were derived from the CAI4 genetic background ( ura3Δ::imm434/ura3Δ::imm434 ) . The DHH1/dhh1Δ heterozygote was generated independently by two approaches: 1 ) by replacement of one endogenous DHH1 allele with a URA3 cassette , and 2 ) by allele replacement using a HIS1 cassette . A transformant generated by each method was tested for filamentous development , and identical results were observed ( Fig 6 and S5 Fig ) . To induce hyphal formation , strains were inoculated onto standard YEPD plates ( 2% glucose , 2% peptone , 1% yeast extract ) supplemented with 80 mg/L uridine and 1% or 10% fetal calf serum ( FCS ) as indicated . Hyphal formation was also induced by growth on carbon-limiting Spider medium ( 10g nutrient broth , 10g mannitol , 2g K2HPO4 per liter media ) [86] .
|
Eukaryotic cells affect precise changes in shape and growth in response to environmental and nutritional stress , enabling cell survival and wild-type function . The single-celled budding yeast provides a striking example , undergoing a set of changes under conditions of nitrogen or glucose limitation resulting in the formation of extended cellular chains or filaments . Related filamentous growth transitions are required for virulence in pathogenic fungi and have been studied extensively; however , the full scope of signaling underlying the filamentous growth transition remains to be determined . Here , we used a combination of genetics and proteomics to identify proteins that undergo phosphorylation dependent upon kinases required for filamentous growth . Within this protein set , we identified novel sites of phosphorylation in the yeast proteome and extensive phosphorylation of mRNA-protein complexes regulating mRNA decay and translation . The data indicate an interrelationship between filamentous growth and these ubiquitously conserved sites of RNA regulation: the RNA-protein complexes are required for the filamentous growth transition , and a well studied filamentous growth signaling kinase is required for wild-type numbers of RNA-protein complexes . This interdependence is previously unappreciated , highlighting an additional level of translational control underlying this complex growth transition .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Large-Scale Analysis of Kinase Signaling in Yeast Pseudohyphal Development Identifies Regulation of Ribonucleoprotein Granules
|
Sensitive , specific and rapid diagnostic tests for the detection of Orientia tsutsugamushi ( O . tsutsugamushi ) and Rickettsia typhi ( R . typhi ) , the causative agents of scrub typhus and murine typhus , respectively , are necessary to accurately and promptly diagnose patients and ensure that they receive proper treatment . Recombinase polymerase amplification ( RPA ) assays using a lateral flow test ( RPA-nfo ) and real-time fluorescent detection ( RPA-exo ) were developed targeting the 47-kDa gene of O . tsutsugamushi or 17 kDa gene of R . typhi . The RPA assay was capable of detecting O . tsutsugamushi or R . typhi at levels comparable to that of the quantitative PCR method . Both the RPA-nfo and RPA-exo methods performed similarly with regards to sensitivity when detecting the 17 kDa gene of R . typhi . On the contrary , RPA-exo performed better than RPA-nfo in detecting the 47 kDa gene of O . tsutsugamushi . The clinical performance of the O . tsutsugamushi RPA assay was evaluated using either human patient samples or infected mouse samples . Eight out of ten PCR confirmed positives were determined positive by RPA , and all PCR confirmed negative samples were negative by RPA . Similar results were obtained for R . typhi spiked patient sera . The assays were able to differentiate O . tsutsugamushi and R . typhi from other phylogenetically related bacteria as well as mouse and human DNA . Furthermore , the RPA-nfo reaction was completed in 20 minutes at 37oC followed by a 10 minute incubation at room temperature for development of an immunochromatographic strip . The RPA-exo reaction was completed in 20 minutes at 39oC . The implementation of a cross contamination proof cassette to detect the RPA-nfo fluorescent amplicons provided an alternative to regular lateral flow detection strips , which are more prone to cross contamination . The RPA assays provide a highly time-efficient , sensitive and specific alternative to other methods for diagnosing scrub typhus or murine typhus .
Rickettsial pathogens are among the leading causes of morbidity and mortality during military operations . In recent years , emerging rickettsial diseases have been reported throughout the world and are a significant medical concern for local and deployed personnel and travelers [1–3] . These pathogens include spotted fever group Rickettsia ( SFGR , Far eastern spotted fever , Japanese spotted fever , Siberian and Queensland tick typhus , and Thai tick typhus just to name a few ) , typhus group Rickettsia ( TGR , epidemic and murine typhus ) and Orientia tsutsugamushi ( scrub typhus , ST ) . Due to the high mortality rate of untreated rickettsial infections , early treatment with appropriate antibiotics is critical [4] . Doxycycline is the drug of choice , except in cases of pregnancy and tetracycline hypersensitivity . Symptoms of rickettsial infections are nonspecific and can be confused with a variety of other pathogens ( e . g . , dengue , malaria , leptospirosis ) that require different treatment regimens . To ensure that appropriate treatment is initiated promptly , early diagnosis of rickettsial infections is critical . Currently , the diagnosis of rickettsial diseases relies mainly on serological methods based on antibody detection [5 , 6] . However , antibody based assays may not be adequate for the diagnosis of disease in the acute phase , as antibody levels may not be detectable at the onset of illness . Therefore , antigen/pathogen detection before the rise of antibody levels is important . Almost all antigen/pathogen detection assays for rickettsial diseases are based on polymerase chain reaction ( PCR [7 , 8] ) , quantitative real-time PCR ( qPCR ) or nested PCR targeting different genes , including 56 kDa [9] , 47kDa [10] , groEL [11] of O . tsutsugamushi and OmpB [7] , 17 kDa [9] , and gltA [12 , 13] of Rickettsia . However , all these assays require end user training to operate the thermocycler , and a functional and well-calibrated thermocycler is difficult to obtain and maintain in resource poor settings . Recently , loop-mediated isothermal amplification has been developed [14–20] as a potential alternative for PCR based tests . In addition to groEL as the target [11] , Huber et al . [21] demonstrated the use of the conserved 47 kDa gene as the target for the LAMP assay , which achieved similar sensitivity to that of qPCR . Recombinase polymerase amplification ( RPA ) uses a mixture of prokaryotic recombinases to guide synthetic oligonucleotide primers to targets in the sample [22 , www . twistdx . co . uk for detail description] . A strand displacing DNA polymerase ( large fragment from Bacillus subtilis Pol I , Bsu ) is used for primer extension [22] . The method is highly sequence specific in complex nucleic acid mixtures and in urine without the need for additional preparation of samples [23] . It offers amplification of the target sequence by reiterative oligonucleotide-primed DNA synthesis without the need to denature DNA at a high temperature . The assay can be performed between 24°C to 45°C [22] with very high efficiency so that detection of product from a single molecule is possible in 20 minutes [22] . The successful application of RPA is evident as shown in many recent publications . RT-RPA has been developed to detect HIV [24] , Rift Valley Fever virus [25 , 26] , Ebola virus , Sudan virus and Marburg virus [26] , MERS-CoV [27] , foot-and-mouth disease virus [28] , and Bovine Coronavirus [29] . Additionally , assays have been developed for detection of Chlamydia trachomatis in urine samples [23] , diagnosis of Cryptosporidiosis in animal and patient specimens [30] , and detection of Neisseria gonorrhoeae , Salmonella enterica , methicillin-resistant Staphylococcus aureus ( MRSA ) [31] , Francisella tularensis [32] , and Group B Streptococci [33] . Furthermore , the assay has been used in combination with ELISA for food analysis [34] . While RPA is similar to LAMP , in that no thermocycler is needed , the optimal reaction temperatures for RPA and LAMP are 37°C and 60°C , respectively . In addition , RPA is potentially easier to adapt to a multiplex format [22] than LAMP as RPA requires only one pair of primers rather than three like LAMP , in which the mixing of multiple pairs of primers may result in non-specific primer interactions and may limit the total amount of primer able to be added to the reaction . Finally , the incorporation of a probe can increase the specificity of the assay . This is particularly important , as the LAMP assay is known to result in non-specific amplification [35] . These advantages suggest that RPA may be a viable point-of-care nucleic acid detection method that can be utilized in resources-limited areas where these diseases are endemic . In this report , we demonstrated that a lateral flow end point detection RPA ( TwistAmp-nfo kit , hereafter RPA-nfo ) could be performed at 37°C in 20 minutes to detect DNA from either O . tsutsugamushi or R . typhi . Similarly , a real-time fluorescent signal detection ( TwistAmp-exo kit , hereafter RPA-exo ) could be performed at 39°C to detect DNA from either O . tsutsugamushi or R . typhi . A detection limit similar to that of the gold standard qPCR was achieved by both RPA-exo and RPA-nfo . The specificity was evaluated using genomic DNA from closely related microorganisms at 1000 folds excess in copy number . The detection limit was established using a range of target genomic DNA from as high as 500 copies per reaction to less than 10 copies per reaction . Further evaluation was done using DNA extracted from liver , lung and spleen collected from O . tsutsugamushi infected mice , DNA extracted from clinical samples obtained from confirmed ST patient sera , or R . typhi-spiked normal human plasma . A cross contamination proof ( XCP ) lateral flow cassette was used to detect RPA-nfo amplicons . This method eliminates the need to open the reaction tube at the end of the reaction , thus minimizing the chance of contamination .
The animal study protocol ( #D11-06 ) was reviewed and approved by the Naval Medical Research Center Institutional Animal Care and Use Committee ( IACUC ) in compliance with all applicable Federal regulations governing the protection of animals in research . The experiments reported herein were conducted in compliance with the Animal Welfare Act and in accordance with the principles set forth in the “Guide for the Care and Use of Laboratory Animals , ” Institute of Laboratory Animals Resources , National Research Council , National Academy Press , 1996 . Oligonucleotide primers used for the RPA assays were manually designed based on the 47 kDa gene sequence from the Karp strain of O . tsutsugamushi ( 47-RPA ) and the 17 kDa gene sequence from the Wilmington strain of R . typhi ( 17-RPA ) per recommendation by TwistDx ( Cambridge , UK ) . A minimum of 5 sets of primers and probes for each target were designed for different end-point detection methods . All primers and probes were synthesized by Eurofins MWG Operon ( Huntsville , AL ) and Biosearch Technologies ( Petaluma , CA ) , respectively . The probes were HPLC purified . RPA-nfo and RPA-exo kits were both purchased from TwistDx ( Cambridge , UK , www . twistdx . co . uk ) . The amplicons generated by the RPA-nfo were detected using a lateral flow strip or cassette ( Fig 1A ) . RPA-exo was used to monitor fluorescent signals in real time using a fluorimeter ( Fig 1A ) . Fig 1B shows a general scheme describing the design of the experiments . PCR product of the 47 kDa gene sequence of O . tsutsugamushi Karp strain and the 17 kDa gene from R . typhi Wilmington strain was cloned into a VR1012 vector and pUC vector , respectively , following standard molecular biology technology . The plasmid was extracted from a 3 ml culture using the Qiagen plasmid mini kit ( Qiagen , CA ) following the manufacturer’s instruction . The concentration was determined using a Nanodrop ( Thermo Fisher Scientific , CA ) and the expected copy number of the target gene was calculated based on the size of the plasmid ( http://cels . uri . edu/gsc/cndna . html ) . The purified plasmid was used as a template to optimize the RPA assays . Genomic DNA from multiple strains of O . tsutsugamushi , including Karp , AFSC4 , AFSC7 , Garton , Ikeda , and Boryong , TGR and SFGR was extracted from renografin gradient purified organisms using QIAamp Mini DNA kit ( Qiagen , CA ) as previously described [36] . Similarly , DNA from cultured Leptospira , Coxiella burnetii , and Bartonella bacillifomis was extracted by the same method . DNA extracted from blood of patients with confirmed ST was provided to us by Dr . Yupin Suputtamongkol of Siriraj Hospital , Mahidol University , Bangkok , Thailand . The animal protocol to perform the mouse experiments ( IDD-11-06 ) was approved by NMRC IACUC . According to the approved protocol , mice were challenged intraperitoneally ( IP ) by the Karp strain of O . tsutsugamushi and observed for up to 21 days . At the indicated time post infection , mice were sacrificed , and various organs were collected which included the lungs , liver and spleen . Genomic DNA was extracted from these organs following the manufacturer’s instructions of the QIAamp Blood and Tissue Mini DNA kit ( Qiagen , CA ) . To prepare cultured R . typhi-spiked NHP , the number of R . typhi present in each of the three independently prepared R . typhi cultures was determined . To do this , DNA was extracted from 100 μl each of the three R . typhi cultures using the QIAamp Blood and Tissue DNA kit following the manufacturer’s instructions . The bacterial load was determined by qPCR to be within the range of 1010 -1011copies/ml . An initial dilution was made by adding cultured R . typhi into NHP at a 1:100 dilution followed by a serial dilution of 1:10 using NHP as a diluent to ensure that the final concentration of the 200 μl of spiked NHP was 2 , 000 copies/ml of cultured R . typhi . The DNA from each 200 μl R . typhi-spiked NHP was extracted using the QIAamp Blood and Tissue DNA kit . The DNA was used as template for the 17-RPA assay to evaluate its performance and qPCR was used to quantitate the copy number of the 17 kDa gene . Each forward primer was mixed with different reverse primers separately to examine which pair performed the best using plasmid DNA ( 10–1000 copies ) as template . Different reaction times , concentrations of primers and probes , and temperatures ranging from 37°C to 42°C were evaluated to select the most sensitive combination for the RPA reactions as per the manufacturer’s recommendations ( TwistDx , Cambridge , UK ) . Each combination of parameters was performed at least twice . The temperature range was varied every 0 . 5 degrees within the range recommended by the manufacturer . The evaluation was done using the RPA-nfo for the RPA reaction and the results were evaluated using Milenia Genline Hybridetect-1 ( MGH ) strips by Millenia Biotec GmbH ( Gieben , Germany ) . Based on the evaluation performed in previous section , a final condition to carry out the recombinase polymerase amplification is described below . The reaction mixture was 50 μl using the RPA-exo or RPA-nfo . The reaction mixture contained rehydration buffer , recombinases , and Bsu strand displacing polymerase with the addition of 420 nM each of the forward and reverse primers , 120 nM of the FAM-tagged probe , at a total volume of 42 . 5 μl . After mixing these components , 2 . 5 μl of 280 mM magnesium acetate was pipetted into the tube lids , and 5 μl of DNA was added to the reaction mixture unless otherwise indicated . The lids were closed and the magnesium acetate was spun down into the reaction mixture to initiate the reaction . For RPA-exo , the tubes were placed into an ESEQuant tube scanner device ( Qiagen , CA ) and the reaction was conducted at 39°C for 20 minutes . The tubes were briefly agitated at 4 minutes after the initiation of the reaction , spun down and placed back into the tube scanner . The reaction in each tube was monitored in real time following the increase of fluorescent signals . For RPA-nfo , tubes were placed into a heating block at 37°C for 20 minutes . The tubes were briefly agitated at 4 minutes after the reaction was initiated , spun down and placed back into the heating block . The MGH strips were used to evaluate results according to the manufacturer’s instruction after the reaction was completed . For each reaction , 1 μl of sample amplicon was used for strip development . As an alternative detection method , after the reaction was completed , the entire reaction tube was inserted into the XCP cassette ( BioUstar , Hongzhou , China ) and the cassette was closed per the manufacturer’s instruction . The presence of a signal at the T line was an indication of positive detection of DNA using a 5’ FAM labeled probe . The copy number of genomic DNA extracted as previously described was determined by qPCR ( below ) . The DNA was diluted to the desired copy number in a total volume of 5 μl to perform the RPA as described previously . The reaction was performed a minimum of 6 times for each amount of genomic DNA . Due to the limited availability of extracted DNA from scrub typhus confirmed patients , only 47-RPA-nfo was performed with 1 μL of DNA . For the DNA from O . tsutsugamushi-infected mice , the DNA was extracted from liver , spleen and lungs of these mice as described previously . The evaluation of 47-RPA-exo and 47-RPA-nfo was performed as described above using 5 μl of extracted DNA . Quantitative PCR was performed to compare and confirm the detection limit of the RPA assay . Whenever the copy number of a sample is provided , it was obtained via qPCR using a standard curve built with a plasmid of known copy number . The 7500 Fast Real-time PCR System ( Applied Biosystems , Foster City , CA ) was used to perform qPCR reactions and analyze the results . The primers used here are the same as those previously used for amplifying the 47kDa gene sequence of the Karp strain of Orientia and 17 kDa gene sequence of R . typhi [10 , 37] . A total reaction mixture of 20 μl contained 1 μM each of forward primer and reverse primer targeting the 47kDa gene of Orientia or 17 kDa gene of R . typhi , 1X RT2 SYBR Green qPCR Mastermix ( SABiosciences , Frederick , MD ) , and DNA template . An initial 5-minute activation step at 95°C was followed by 40 cycles of 10 seconds at 95°C , 30 seconds at 60°C , and a melting curve determination cycle . To determine the amount of mouse genomic DNA in the total DNA extracted from organs of O . tsutsugamushi infected mice , qPCR was performed as described by Sunyakumthorn et al [38] .
After running the RPA-nfo reaction with different primer and probe combinations , incubation temperatures and reaction time as mentioned in Materials and Methods , we applied several performance factors including total signal strength , absence of background signal on the MGH strip and detection limit to determine which combination of primer and probe , including concentration , reaction temperature and reaction time provided the lowest limit of detection . We used qPCR as a gold standard to quantitate the DNA to establish the detection limit . A primer and probe set that provided the best detection limit for the individual detection of O . tsutsugamushi or R . typhi DNA was identified ( Table 1 ) . Fig 2A shows typical results for the 47-RPA-nfo detection of different copy numbers of the 47 kDa gene . The 47-RPA-nfo appeared to be detecting 53 copies/reaction which is slightly higher than that by qPCR ( 10 copies/reaction ) . Different strains and isolates of O . tsutsugamushi including Karp , Kato , Gilliam , AFSC4 , AFSC7 , Ikeda , Garton , and Boryong were tested . While the region ( 156 bp ) of the 47 kDa sequence that RPA primers targeted shared greater than 90% homology among these 8 strains ( S1 Fig ) , the detection limit ranged from 10 to 400 copies per reaction . Similarly , the detection of R . typhi DNA using different copy numbers of the 17 kDa gene is representatively shown in Fig 2B . The 17-RPA-nfo was able to detect around 20 copies per reaction which is slightly higher than that by qPCR ( 6 copies per reaction ) . Results shown in Fig 3A indicate that the level of detection of the 47-RPA-exo was around 50 copies per reaction , this is similar to that of 47-RPA-nfo and qPCR at 53 and 10 copies per reaction , respectively . Similarly , the 17-RPA-exo showed a level of detection of 40 copies per reaction ( Fig 3B ) comparable to that of 17-RPA-nfo and qPCR , at 20 copies and 6 copies per reaction , respectively . Furthermore , the significant rise of fluorescent signal over time ( i . e . slope ) was observed no later than 10 minutes post initiation of the reaction , even though the signals plateaued within 20 minutes , suggesting the possibility of differentiating positive and negative samples based on the slope which reflects the change of fluorescent signal with time . To further evaluate whether the 47-RPA or 17-RPA only detected O . tsutsugamushi or R . typhi , respectively , purified genomic DNA from phylogenetically related organisms was tested . When DNA from R . typhi , R . bellii , R . rickettsii , R . conorii , Leptospira , C . burnetii , B . bacilliformis was used in the 47-RPA-exo , none were positive ( Fig 4A ) . Consistent with this observation , the 17-RPA-exo did not show any reactivity with DNA from O . tsutsugamushi , Leptospira , C . burnetii or B . bacilliformis ( Fig 4B ) . The same results were obtained using 47-RPA-nfo . It is noted that when 17-RPA-nfo was used to detect R . conorii and R . rickettsii , positive results were only observed when at least 104 copies/reaction of DNA were present ( S2 Fig ) , suggesting that the assay was much less sensitive to detect these two SFGR . Other SFGR , such as R . honei and R japonica were also evaluated to show that they were not detectable by 17-RPA-nfo unless 104 or more copies of DNA were present , same results were obtained using 17-RPA-exo . Additionally , the performance of 17-RPA-exo was also evaluated using DNA extracted from R . prowazekii . The assay consistently detected as low as 80 copies per reaction . Taken together , these results are consistent with the notion that this 17-RPA is better suited to detect TGR . Different amounts of extracted genomic DNA were used to evaluate the detection limit of the two RPA methods running at least 6 replicates for each assay . The 47-RPA-nfo and 47-RPA-exo had different detection limits whereas the 17-RPA-nfo and 17-RPA-exo had very similar detection limits . As shown in Table 2 , when testing 47-RPA-exo , the assay detected 100% , 77% and 42% of samples with 100–120 copies , 40–60 copies and 10–12 copies per reaction , respectively . When 47-RPA-nfo was used , the assay was positive for 100% , 67% , 54% and 0% of samples with 500–550 copies , 200–250 copies , 100–120 copies and 40–60 copies per reaction , respectively . Based on these results , 47-RPA-exo had a significantly lower detection limit from that of 47-RPA-nfo . The 17- RPA-exo detected 88% , 91% and 43% of samples ranging from 100–120 copies , 40–60 copies and 5–15 copies per reaction , respectively . The 17-RPA-nfo , on the other hand , detected 100% , 73% , 71% and 57% of samples ranging from 100–120 copies , 40–60 copies , 20–25 copies and 5–15 copies per reaction , respectively . There was no statistical significant difference ( Table 2 , Fisher’s exact test ) in percentage of positive samples detected for various ranges using 17-RPA-exo and 17-RPA-nfo , suggesting that the detection limit was comparable between 17-RPA-exo and 17-RPA-nfo . The 47-RPA was evaluated using a total of 10 positive and 10 negative human samples , the results showed that 8 out of 10 positives were 47-RPA-nfo positive while all 10 negatives were negative ( Fig 5 ) , similar to what was observed using the LAMP method [21] . Even though we only ran 47-RPA-nfo using 1 μl instead of 5 μl DNA due to the limited availability of extracted DNA , nevertheless , the limited number of clinically confirmed ST patients shows that 47-RPA exhibited 80% sensitivity and 100% specificity . The 47-RPA was also evaluated using extracted DNA from liver , spleen and lung from laboratory , O . tsutsugamushi-infected mice . These samples were confirmed qPCR negative before day 4 post infection and then a gradual increase in quantity was seen as shown in Table 3 . The same extracted DNA was used as a template for both 47-RPA-exo and 47-RPA-nfo . Among these samples , 3 qPCR negative samples were determined negative by 47-RPA-exo and 47-RPA-nfo . Among the 9 qPCR positive samples , 7 of them were detected positive by 47-RPA-exo and 47-RPA-nfo . These results demonstrated that 78% sensitivity and 100% specificity was achieved using DNA extracted from organs obtained from O . tsutsugamushi-infected mice . Taken together , it is concluded that 47-RPA provides around 80% sensitivity and 100% specificity . Due to the lack of qPCR confirmed R . typhi positive patient samples , the clinical performance of the 17-RPA was evaluated by spiking normal human plasma with cultured R . typhi . The results of the qPCR quantitation , 17-RPA-exo , and 17-RPA-nfo are shown in Table 4 . All qPCR positive samples were positive by 17-RPA-exo and 17-RPA-nfo . Based on the amount of R . typhi DNA spiked into NHP and the final copies of R . typhi in the extracted DNA , the DNA extraction resulted in average DNA recovery of 81% ( 48%–104% ) . Using the XCP cassette to replace the MGH strips at the end of the RPA-nfo reaction , the FAM labeled probes showed signal at the T line ( Fig 6 ) , just like the strips . These results demonstrated the potential usage of this XCP cassette as a replacement for the MGH strips for the detection of RPA-nfo amplicons . As the XCP cassette does not require opening the reaction tube and uses all of the reaction volume for detection , it is conceivable that the usage of the XCP cassette to detect the end product could be more sensitive , and present less opportunity for contamination than the usage of MGH strips to detect the RPA-nfo product .
The data presented here shows that RPA is a method that could be used to detect plasmid DNA or DNA extracted from patient samples , infected mice , and pure organisms with a detection limit of tens of copies , within 20 minutes . This is generally comparable to that of qPCR ( Figs 1 and 2 and Table 3 ) . This detection limit observed is not likely due to the limitation of the RPA assay as it has been shown to detect product amplified from a single molecule [22] . Rather , it is conceivable that the detection limit can be improved if different genes are selected for assay development . In addition to having a similar detection limit , the RPA method has several advantages over qPCR , making it an attractive alternative . Firstly , a heating block that is capable of maintaining a temperature of 37°–39°C for 20 minutes is sufficient to perform the reaction . Secondly , reaction mixtures are pre-made and provided by TwistDx . With the addition of water , template , primers and probe , the reaction is then initiated upon mixing , thus minimizing the potential for contamination often observed with other nucleic acid amplification methods . Thirdly , the method offers multiple end-point detection options with similar detection limits that could fit into many different laboratory settings , thus making one assay applicable in a well-equipped laboratory , a mobile laboratory or a rural area where instruments and infrastructure may be limited . Finally , successful detection of O . tsutsugamushi from clinical and mouse samples known to be infected , demonstrated the potential clinical application of the assay . The RPA method was developed almost 10 years ago . However , it has not been accepted as well as some other isothermal amplification methods , i . e . , LAMP . Although the method is relatively easy to perform with all the advantages described previously , one difficulty lies in the design of the primers and probes . There is no software available to assist in properly designing primer and probe sets . Furthermore , the requirement of extra length primers and probes has made the manual design more difficult than other methods , such as qPCR , PCR and LAMP which have well established software for primer and probe design . The relatively rare modifications on the probes also add cost to the synthesis . Consequently , it may prevent the evaluation of large numbers of potential primers and probes to lead to the most sensitive and specific assay possible . Nevertheless , the advantages seem to outweigh these minor drawbacks as we and many others [22–34] were able to demonstrate the utility of the assay for detection of closely related pathogens without sacrificing sensitivity or specificity . In addition to the development of RPA for detection of O . tsutsugamushi and R . typhi , the RPA-nfo method was developed C . burnetii DNA [39] . The detection limit of these assays was comparable to that of qPCR . In the detection of O . tsutsugamushi using RPA methods , there appeared to be a slight difference in detection limit between RPA-nfo ( 54% of 100–120 copies per reaction ) and RPA-exo ( 77% of 40–60 copies per reaction ) . On the contrary , the difference in detection limit was less pronounced for R . typhi detection as shown in Table 2 . Due to the observations that RPA-nfo requires higher copies of target gene for positive detection in general , the RPA-nfo assays for both targets were evaluated using higher copies/reaction or additional ranges of copies/reaction to best estimate the detection limit of RPA-nfo assay . The test results using varying ranges of genomic DNA for 47-RPA and 17-RPA-demonstrated differences in percent positive detection , for most tested ranges , the assay provided around 50% positivity regardless of whether it was RPA-nfo or RPA-exo . Additionally , these results are consistent with the notion that lateral flow strips are generally regarded as not-as-sensitive as other detection methods , such as fluorescent signal detection . The use of strips to evaluate the results of RPA-nfo requires the reaction tubes to be opened to remove samples after the reaction is completed , thus adding extra post reaction time and increasing the possibility of cross contamination due to the presence of abundant amplicons . The cassette form of the lateral flow strips essentially eliminates cross contamination yet still requires additional time to obtain the final results . It is worth noting that both MGH strips and XCP cassettes are commercially available and they were used without any attempt to optimize the performance . Thus , it is conceivable that further optimization of these devices could improve the sensitivity and eliminate the difference in sensitivity between 47-RPA-nfo and 47-RPA-exo . The target genes , 47 kDa and 17 kDa , were selected for O . tsutsugamushi and R . typhi DNA detection , respectively , based on their relatively conserved nature in sequence . The 47 kDa gene codes one of the major protein antigens recognized by sera from patients infected with many strains of O . tsutsugamushi [40] . The sequences of 47 kDa proteins have been compared to show greater than 96% identity [41] between strains with the exception of one recently identified strain , O . chuto [42] , that shared only about 83% identity [41] . The gene has been used as the target for the development of PCR , qPCR , and LAMP assays and has shown consistent results in sensitivity , specificity , and broad reactivity toward many different strains of O . tsutsugamushi [10 , 21] . While the region ( 156 bp ) of 47 kDa that RPA targeted shared greater than 90% homology among the 8 strains tested ( S1 Fig ) , it is noted that a different detection limit was observed between strains . In spite of the difference in detection limit observed for different strains , these detection limits still fall within the broad expected level of O . tsutsugamushi circulating in patients’ blood [43] . Since the level of O . tsutsugamushi present in patient’s blood is also associated with the disease severity , it is likely that the current 47-RPA-exo and 47-RPA–nfo need further improvement to achieve lower detection limits in order to be more clinically applicable . Similarly , the 17 kDa gene has been used to confirm the presence of Rickettsia DNA in potentially infected samples [9 , 37] . Additional targets were used to delineate whether the infection was due to TGR or SFGR [7 , 8] . It is noted that the designed primers and probe for 17-RPA , which are located at the latter half of the 17 kDa sequence , showed a lower detection limit for R . typhi DNA while it required a copy number greater than 104 of R . conorii and R . rickettsii DNA to be detectable by 17-RPA-nfo ( S2 Fig ) . Same observation was made for R . honei and R . japonica . The 17 kDa genes from SFGR that were tested share 89% of the identity with the full length 17 kDa gene of R . typhi . The 17 kDa gene of R . typhi , shares 95% of the identity with the 17 kDa gene of R . prowazekii , another member of the TGR . The 17 kDa gene of R . conorii , R . rickettsia , R honei , and R . japonica show no more than 87% of the identity with that of R . typhi within the region where the primers and probe are located ( 134 bp ) , the identity between R . typhi and R . prowazekii is 97% . Since the sensitivity is clearly different between R . typhi and the two species in the SFG Rickettsia using just one set of primers and probe , it suggests that 87% identity is not enough to use one set of primers and probe to detect both TG and SFG Rickettsia with equal sensitivity . Therefore , another set of primers and probe is needed for SFGR in order to have a similar detection limit to that for R . typhi . This difference in reactivity toward TGR and SFGR suggests that the assay is highly specific . While the detection limit using NHP spiked with cultured R . typhi was about 2000 copies/ml ( Table 4 ) , it is higher than the median of 210 DNA copies/mL in blood of confirmed murine typhus patients [44] , suggesting that the assay may not be sensitive enough to be clinical useful . Since no evaluation of 17-RPA was done using clinical samples , it is warranted that additional evaluations should be done to further characterize the clinical sensitivity and specificity of the 17-RPA . Both RPA-exo and RPA-nfo kits are specific only to the targeted organisms . This is demonstrated in Fig 3A and 3B , and Table 3 . In Fig 3 , it is shown there was no amplicon unless the target DNA ( i . e . , O . tsutsugamushi or R . typhi ) was present even though the non-target DNA was present in 1000 folds excess compared to the target gene . The presence of excessive human DNA and mouse DNA apparently did not interfere with the detection of the target gene , suggesting that the assay provided sufficient specificity for clinically relevant samples . This is supported by the results showing that 8 out of 10 PCR confirmed O . tsutsugamushi positives were positive and 10 out of 10 PCR confirmed O . tsutsugamushi negatives were negative . While we achieved 80% sensitivity and 100% specificity using a limited number of clinical samples , we also evaluated the specificity and sensitivity of the 47-RPA-exo and–nfo more closely using samples from mice infected by live O . tsutsugamushi . As shown in Table 3 , no samples collected prior to day four post infection had qPCR detectable O . tsutsugamushi . As expected due to the similarity in sensitivity between qPCR and RPA , none of these samples were positive by either 47-RPA-nfo or 47-RPA-exo . On the contrary , as the infection progressed , O . tsutsugamushi DNA became detectable by qPCR and 47-RPA as early as day four post infection in all organs evaluated , even though the number of O . tsutsugamushi detected was low . For those samples with low copy of O . tsutsugamushi DNA by qPCR , not all of them were 47-RPA detectable . Overall , both 47-RPA-exo and–nfo showed consistent results of 78% sensitivity and 100% specificity , similar to the observation made using DNA extracted from scrub typhus patients . It is noted that the number of O . tsutsugamushi was undetectable before day 4 post infection or the number was very low at day 4 post infection . This is probably related to the number of live organisms injected into the mice , the naturally-slow growth of O . tsutsugamushi and the route and rate of dissemination of O . tsutsugamushi once it enters into the mice . The amount of mouse DNA was not the same for all samples , so we normalized the copy number of O . tsutsugamushi to a mouse complement factor D gene [cfd , 38] . The ratio of mouse DNA to O . tsutsugamushi DNA ranged from as low as 1 . 9 ( lung , day 11 ) to as high as 76800 ( lung , day 4 ) . These results suggested that the ratio of non-target DNA to target DNA did not affect the detection of target DNA with a 104 dynamic range and that either 47-RPA-nfo or 47-RPA-exo has the potential applicability in clinical samples for diagnosis of infection . Nevertheless , the number of samples evaluated in this report is limited and the evaluation of additional clinical samples is needed to better describe the clinical sensitivity and specificity . In this work , we took advantage of the similarity between LAMP and RPA-nfo in detecting FAM labeled amplicons , provided a reverse primer is labeled with biotin . The BioUStar XCP cassette was designed specifically for the detection of LAMP amplicons when a FAM labeled loop primer along with a biotin labeled reverse primer were used together in the reaction . In the original design , the C-line signal indicates an internal control containing a DIG labeled loop primer while the T-line signal indicates a positive sample containing a FAM labeled loop primer ( www . bioustar . com ) . The results in Fig 6 clearly demonstrated that the assay was only positive at the T line when correspondingly labeled probe was present together with all primers and DNA templates . While it is true that the XCP cassette was not designed to detect RPA-nfo products , our approach confirmed the possibility of using the XCP cassette for RPA-nfo to provide an end point readout with no cross contamination . Furthermore , this provides a single XCP cassette that can be used for the detection of practically all FAM ( or DIG ) labeled probe amplicons , including qPCR , LAMP and RPA-nfo in a pick-and-choose manner to select those assays for target organisms that are most relevant to the local area where the assay will be performed . In conclusion , the work presented here demonstrates the development of RPA-nfo and RPA-exo for the detection of O . tsutsugamushi or R . typhi DNA . The assay had a detection limit similar to that of qPCR . The specificity of the assays was evaluated using excessive amounts of mouse DNA and this did not affect the reaction . The assays were also evaluated using extracted DNA from human patient samples , demonstrating around 80% sensitivity and 100% specificity using limited clinical samples . Finally , the ease with which the cross-contamination-proof lateral flow cassette can be used to detect multiple amplicons from various nucleic acid amplification methods makes it promising for wide-ranging use in the field .
|
Historically , rickettsial pathogens are among the leading causes of morbidity and mortality during military operations . Rickettsial diseases , lately , are reemerging in areas of known abundance or emerging in areas of unknown existence , posing a significant medical concern for local residents and travelers . The diseases are difficult to diagnose as they often share similar symptoms with many other diseases in the same geographical areas . Therefore , it is particularly challenging for clinicians to provide a timely and accurate diagnosis . A recombinase polymerase amplification ( RPA ) -based nucleic acid detection platform has been used to develop accurate , sensitive , specific , and easy-to-perform assays to detect O . tsutsugamushi or R . typhi , indicative of scrub typhus or murine typhus , respectively . These RPA assays provide similar limits of detection and specificity to that of qPCR . Unlike qPCR , they require no thermocycler and provide multiple end-point monitoring options amendable to different laboratory capabilities . This work presents an alternative assay platform for early detection of O . tsutsugamushi or R . typhi infection so that timely treatment can be prescribed in well-equipped laboratories as well as resource limited areas .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
Development of Recombinase Polymerase Amplification Assays for Detection of Orientia tsutsugamushi or Rickettsia typhi
|
Compositional data consist of vectors of proportions normalized to a constant sum from a basis of unobserved counts . The sum constraint makes inference on correlations between unconstrained features challenging due to the information loss from normalization . However , such correlations are of long-standing interest in fields including ecology . We propose a novel Bayesian framework ( BAnOCC: Bayesian Analysis of Compositional Covariance ) to estimate a sparse precision matrix through a LASSO prior . The resulting posterior , generated by MCMC sampling , allows uncertainty quantification of any function of the precision matrix , including the correlation matrix . We also use a first-order Taylor expansion to approximate the transformation from the unobserved counts to the composition in order to investigate what characteristics of the unobserved counts can make the correlations more or less difficult to infer . On simulated datasets , we show that BAnOCC infers the true network as well as previous methods while offering the advantage of posterior inference . Larger and more realistic simulated datasets further showed that BAnOCC performs well as measured by type I and type II error rates . Finally , we apply BAnOCC to a microbial ecology dataset from the Human Microbiome Project , which in addition to reproducing established ecological results revealed unique , competition-based roles for Proteobacteria in multiple distinct habitats .
A long-standing goal of applied statistics in many fields has been identifying features associated significantly by a measure such as correlation [1 , 2] . When the features to be associated form a composition , inference of the correlation matrix is subject to the well-known problem of spurious correlation [3–6] . Compositional data in particular are vectors of proportions that sum to a fixed constant ( typically one ) ; they are usually thought of as the result of sum-normalizing an unobserved ( or unrecorded ) and unconstrained basis , following the terminology of [6] . The resulting sum-constraint of the compositional data means that any pairwise correlation measured using such data can be non-zero even if all the pairwise correlations on the unobserved count scale are zero , a phenomenon called spurious correlation [3] . The fact that all the features sum to one also makes the correlation matrix on the unobserved counts ( that is , the basis correlation matrix ) non-identifiable without untestable , though perhaps not unreasonable , assumptions [7–10] . Any method thus offers at best a partial reconstruction of the unobserved count correlation matrix , and the interest in characterizing such correlations in fields from geology to ecology has led to a variety of approaches . In the context of microbial ecology , several methods have been proposed to identify significant ecological relationships from compositions; virtually all rely on some form of sparsity assumption and infer quantities relating to the log-transformed unobserved counts ( hereafter referred to as the log-basis ) . The only technique that does not rely on a sparsity assumption is ReBoot [7] , which estimates a “compositionally-corrected” correlation matrix using a permutation-based method . Friedman and Alm [8] proposed SparCC , which estimates the log-basis correlation matrix under the assumption that the correlations are on average small in magnitude . Fang et al . [9] noted that the resulting estimate is not guaranteed to be positive definite or that the elements will lie inside [–1 , 1] and proposed CCLasso to estimate the log-basis correlation matrix using a LASSO penalty on the off-diagonal elements of the variance-covariance matrix . Ban et al . [10] similarly proposed REBACCA to estimate the log-basis correlation matrix; they use the same LASSO penalty function but a different likelihood function . Kurtz et al . [11] proposed SPIEC-EASI to estimate the log-basis precision matrix when the number of features is large by using sparse graph estimation techniques . These approaches have difficulty quantifying uncertainty in the estimates , cannot incorporate uncertainty from the choice of tuning parameter , and are not flexible in the quantities they estimate . Friedman and Alm [8] proposed an inferential procedure based on the bootstrap , but offered no theoretical justification . Fang et al . [9] and Kurtz et al . [11] focused solely on estimation , while Ban et al . [10] used a subsampling method from Shah and Samworth [12] to stabilize the selection error rate . The LASSO-based methods [9–11] typically choose a shrinkage parameter and subsequently infer the log-basis covariance or precision matrix . Friedman and Alm [8] , Fang et al . [9] , and Ban et al . [10] all use the log-basis covariance matrix for network construction , while Kurtz et al . [11] use the log-basis precision matrix . This means that investigators typically must choose whether a precision or correlation matrix is best , and often use the resulting estimate with little guidance as to its uncertainty . We address these issues by providing a flexible , fully Bayesian approach to identify correlations in compositional data . It is able to quantify uncertainty through the associated posterior and estimates both the log-basis correlation and precision matrix by modeling the composition directly . The graphical LASSO prior of [13] is used to estimate a sparse log-basis precision matrix ( and hence a sparse log-basis correlation matrix ) through a LASSO penalty , mitigating the non-identifiable nature of the unobserved count correlation matrix . We have implemented the resulting method as BAnOCC ( Bayesian Analysis of Compositional Covariance ) . In this study , we also use a first-order Taylor expansion to approximate the compositional covariance as a function of the mean and variance of the unobserved counts . While not necessary to the development of our method , this expansion helps us explore the situations in which a naïve approach ( ignoring the sum-constraint ) might work . This approximation shows not only that the spurious correlation between two features can take any value in [−1 , 1] even if none of the features are correlated on the unobserved count scale , but also that both the variances and means of the unobserved counts control the magnitude and direction of the spurious correlation . Thus , we provide a novel characterization of the surprisingly broad circumstances under which compositionality can impede straightforward identification of the correlation matrix , and we provide the BAnOCC model to overcome this in datasets where it is possible .
The model assumes that a single subject’s composition , Ci = ( Ci , 1 , … , Ci , p ) T , is generated by the normalization of that subject’s unobserved and unconstrained counts , Xi = ( Xi , 1 , … , Xi , p ) T . That is , Ci=Xi∑j=1pXi , j . We also assume that the unobserved counts for all subjects are independent and identically distributed ( iid ) ; this implies that the compositions are iid as well because the transformation is per-subject . We also introduce notation for the covariance and correlation among the features . The covariance matrix of the unobserved counts is denoted by ΣX = [σX , jk] , to be inferred from C1 , … , Cn . Similarly , the covariance matrix of the composition is denoted by ΣC = [σC , jk] . To construct the network of feature interactions , the relevant null hypotheses ( one for each feature pair j and k ) are that features j and k have a covariance of zero ( σX , jk = 0 ) ; this is equivalent to testing if they are uncorrelated ( ρX , jk = 0 ) . We then define the unobserved count and compositional correlation matrices as RX = [ρX , jk] and RC = [ρC , jk] , respectively .
The likelihood is parametrized by the log-basis precision matrix O = S−1 and the log-basis mean m , and other parameters of interest like the log-basis covariance matrix S are sampled as transformations of these . By parametrizing using O , we are able to leverage a graphical LASSO prior to enforce sparsity on O and by extension S . Conveniently , the assumption of the log-normal distribution obviates the need to sample the covariance of the unobserved counts to determine the existence and direction of an association between two features on the unobserved count scale . This results because when some element of S , sjk , is zero , then the corresponding element of ΣX , σX , jk∝esjk−1 will also be zero; further , the non-zero elements of S and ΣX will have the same sign ( though not the same magnitude ) . Under the log-normal assumption , the complete likelihood of the observed composition ci and the latent total ti=∑j=1pxi , j is given by L ( m , O|ci , ti ) =exp[−12{log ( citi ) −m}TO{log ( citi ) −m}] ( ti ) ( 2π ) p/2|O|−1/2∏j=1pci , j , ( 1 ) where ci= ( ci , 1 , … , ci , p−1 , 1−∑j=1p−1ci , j ) . A detailed derivation can be found in S1 Text . Fitting this likelihood directly is computationally expensive , as the presence of the latent totals necessitates exploring a space whose dimension depends on both n and p . However , ( 1 ) factors into two portions: a part dependent on the compositions ci , and the kernel of a log-normal distribution for the totals ti=∑j=1pxi , j with parameters mi*=1TO ( m−log{ci} ) s2* and s2*=11TO1 ( where 1 is a vector of 1’s ) . Integrating over the totals in ( 1 ) ( S1 Text ) gives the more computationally tractable marginal likelihood L ( m , O|ci ) =| O|1/2exp{ −12 ( m− logci ) TO ( m−logci ) − ( m*i ) 2S2*} ( 2π ) 1/2 ( s2* ) 1/2 ( 2π ) p/2Πj=1pcij . In order to mitigate the non-identifiability of the precision matrix O , BAnOCC uses a shrinkage prior to conservatively estimate the sparsest O consistent with the observed relative abundance data . This is the graphical LASSO prior of [13]: p ( O|λ ) =C-1∏j=1pExpojj|λ2{∏k=i+1pLaplaceojk|λ}1O∈M+ , where 1O∈M+ is an indicator function that O is positive definite , Exp ( x|λ ) has the exponential density of the form p ( x ) = λe−λx1x>0 , and Laplace ( x|λ ) has the Laplace density of the form p ( x ) =λ2e−λ|x| . In comparison to variable selection priors such as spike-and-slab [18] , the graphical LASSO prior is more scalable to high dimensions at the cost of being unable to generate estimates that are exactly zero [19] . We deal with this by using the resulting posterior samples to conclude whether a correlation is likely to be zero or not . The choice of λ is key to the degree of shrinkage imposed by this prior . We placed a gamma prior on λ in lieu of specifying it a priori; this is possible because [13] showed that the normalizing constant C does not depend on λ . The prior for m is the conditionally-conjugate normal prior N ( n , L ) with mean n and covariance matrix L . Hyperparameter choice for the two priors ( on m and λ ) is discussed in more detail below . BAnOCC samples the posterior using Stan’s C++ implementation and R interface [20] . Multiple quantities can be estimated from BAnOCC , including the log-basis precision , covariance , and correlation matrices . In our simulations and application , we estimated the log-basis correlation RlogX because it is interpretable and nicely scaled; we used the posterior median as the point estimate and the 95% credible intervals for wjk to determine whether the correlation between features j and k was non-zero .
The interpretation of the prior parameters on m is relatively straightforward , while that of the shrinkage parameter λ is less clear . Because log-basis means m have a normal distribution , em represents the median unobserved counts , which conveniently have a log-normal distribution with parameters n and L . Therefore , we could parametrize the prior on m by the expected median unobserved counts nLN = exp{n + 0 . 5diag ( L ) } and uncertainty of the median unobserved counts LLN=nLNnLNT ( eL−1 ) . The prior on the shrinkage parameter λ has a shape parameter a that determines how much prior probability mass is placed on λ values close to zero , and a rate parameter b that determines how the probability mass is spread across the entire domain . In particular , a ≤ 1 forces an asymptote at zero , while a > 1 does not . When little or no prior data is available , weakly informative priors can be used . Any prior on λ should have high probability mass close to zero and so should have a ≤ 1 . Larger values of a will “soften” the asymptotic behavior at zero ( S1 Fig ) . The value of the rate parameter b should be chosen to so that most prior probability mass is on sensible values for λ . The degree of shrinkage implied by λ does not appreciably change for λ > 1 ( S2 Fig ) , and so a b of around 5 will give a reasonable uninformative prior distribution for λ . For the log-basis means , m∼N ( 0 , lI ) can be used , with l a large value such as 100 . An overlarge value for l can make computation less efficient and put prior mass on grossly implausible values of em , so an l of 500 or less is reasonable . Prior subject-matter information can be incorporated into the priors for both λ and m , but most easily into the prior on m . If the data have few features , a smaller shape hyperparameter a should be employed to upweight values of λ that yield high shrinkage . The implied prior on the median unobserved counts em could be sampled to provide an empirical distribution of the total counts ∑j=1pemj; this could be assessed for gross deviations from what might be considered reasonable , or agreement with known ranges if such data are available .
The implementation of BAnOCC is publicly available with source code , documentation , and tutorial data as an R/Bioconductor package at http://huttenhower . sph . harvard . edu/banocc .
We first aimed to identify what characteristics of compositional data impede or facilitate the accurate estimation of the unobserved count correlation matrices in general . Such characteristics should delineate when BAnOCC or any other technique for estimating the unobserved count correlation would perform well . A first-order Taylor expansion approximates the compositional covariance as a function of the mean and covariance of the unobserved counts . Because the compositional correlation is a function of the compositional covariance , the resulting approximation also explains how the correlation behaves . Letting X represent the unobserved counts and C the composition , with the mean of X denoted by μX = ( μX , j ) T and the approximate average proportions by ω= ( μX , 1∑j=1pμX , j , … , μX , p∑j=1pμX , j ) T , the Taylor expansion yields ΣC≈ ( 1∑j=1pμX , j ) 2 ( I−ω1T ) ΣX ( I−ω1T ) T . ( 2 ) Here I is the p × p identity matrix , and 1 is a p-dimensional vector of 1’s . Eq ( 2 ) allows us to approximate the behavior of the compositional covariance from the parameters of the unobserved counts that generate it . For a detailed derivation , see S1 Text . Surprisingly , when no features are correlated on the unobserved count scale , the spurious correlation can take any value in [−1 , 1] depending on the properties of the unobserved counts ( Fig 2 ) . This is suggested by considering Eq ( 2 ) when σX , jk = 0 for all j ≠ k , then σC , jk≈ ( 1∑l=1pμX , l ) 2[ωjωk∑lσX , ll−ωjσX , kk−ωkσX , jj]forj≠k . ( 3 ) The weights ωj and the variances σX , ll can be configured arbitrarily to force σC , jk either to the extreme positive or extreme negative end of the spectrum . In particular , we see three types of strong spurious correlations ( Fig 2B–2D ) : “negative dominant” , “positive dominant” , and “negative mixed” . These three types of correlations are thus representative of a range of expected real-world behaviors , and we included them in subsequent simulation studies of BAnOCC and previous models . “Negative dominant” spurious correlation ( Fig 2B ) occurs when features j and k in the unobserved counts have ( 1 ) high mean and ( 2 ) high variability compared to the remaining ( l ≠ j , k ) features . Intuitively , the remaining features must contribute minimally to the total mean or total variance in the unobserved counts . When normalized , the sum-constraint thus forces a negative correlation between features j and k because they behave as if they were the only two features in the composition . In the “positive dominant” spurious correlation type ( Fig 2C ) , features j and k in the unobserved counts have ( 1 ) small variability and ( 2 ) high mean relative to the remaining ( l ≠ j , k ) features . The positive correlation in the composition results because the variability in the sum of the remaining feature abundances causes the compositions for features j and k to be shrunk or stretched in the same direction when the data are normalized . Finally , “negative mixed” spurious correlations are the result of “positive dominant” type bases where feature k and the remaining features have switched roles ( Fig 2D ) . After normalization , the variability in feature k forces feature j to move in the opposite direction to accommodate the remaining features . Eq ( 3 ) also offers an alternative explanation for the negative covariance between features in a Dirichlet distribution . A Dirichlet distribution with parameters α1… , αp results when each feature is independent on the unobserved count scale and has a Gamma ( αj , β ) . The mean and variance of a Gamma distribution are αβ and αβ2 , respectively , implying that in the unobserved counts , a feature with high mean will also have high variance , and vice versa . This captures “negative dominant” correlations well , but fails to capture “positive dominant” or “negative mixed” correlations , which result when at least one feature has high mean but low variance in the unobserved counts . Eqs ( 2 ) and ( 3 ) further suggest that the overall effect of normalization on the correlation estimate as the number of features p increases depends on the characteristics of μX and ΣX . In ecological applications , it is often assumed that if p is large and the compositional means are similar across the p features , then the correlation estimates based on the composition and unobserved counts are not likely to be very different [8 , 10] . Part of the appeal of this reasoning is that it does not rely on information about the unobserved and unconstrained counts . Expanding Eq ( 2 ) , we can see that ΣC ∝ ΣX − ω1TΣX − ΣX1ωT + ω1TΣX1ωT . If the means are very similar to each other , this affects only the weights ω given to the offset ω1TΣX − ΣX1ωT + ω1TΣX1ωT . Small weights render the offset negligible only in the case where the unobserved variance on the unobserved counts ΣX is not too large: the behavior of the offset as the number of features increases depends on the similarity of the means ( through ω ) and on the variances of the additional features in the unobserved counts ( through ΣX ) . Thus when analyzing compositional data , one cannot know with certainty in which data the correlations are strongly affected by the normalization , much less the magnitude and direction of the change in correlation structure induced by normalizing . The information loss due to normalization implies that ΣX is non-identifiable without assumptions about its structure . However , knowing how the unobserved and unobserved counts affect the spurious correlation allows simulation of datasets that have specific types of spurious correlation for testing the performance of estimation methods in these cases .
Using the information from this theoretical analysis , we tested BAnOCC on two types of datasets . The first comprised small datasets generated using the model itself but designed to be challenging by incorporating negative dominant correlations . Second , we also simulated larger , more realistic datasets using an independent model specific to microbial community structure , sparseDOSSA [21] . For the former , four small datasets with 1 , 000 samples and nine features each were generated according to four scenarios . The “simple” scenario had no true correlations and no negative dominant correlation; the “high spurious” scenario had no true correlations but the presence of a negative dominant correlation; the “retained spike” scenario had several true correlations and no negative dominant correlation; and the “reversed spike” scenario had several true correlations and a negative dominant correlation between two features that are positively correlated in the unobserved abundances ( see details in S2 Text and data in S1 Data ) . On these data , we used hyperparameters nj = 0 , L = 1000I , a = 0 . 5 and b = 5 ( S3 Fig ) . Realistic data were generated using the SparseDOSSA model [21] , which generates each feature from a zero-inflated , truncated log-normal distribution with subsequent rounding and estimates the feature-specific parameters by fitting to a given real-world template dataset . We induced correlations between features by using a multivariate distribution with a log-basis correlation that had off-diagonal elements set to one of four different correlation strengths ( {−0 . 7 , −0 . 3 , 0 . 3 , 0 . 7} ) . To ensure that strong compositional effects were present , we used a template with low-diversity community structure [22] with 14 pseudomicrobial features . The correlations were set so that the non-zero elements of the log-basis precision matrix and the log-basis covariance matrix would be the same; we used seven correlations ( see details in S2 Text and data in S2 Data ) . We used hyperparameters a = 0 . 5 , b = 5 , nj = 3 , and L = 30I ( S4 Fig ) . Using our first set of simulated data for evaluation , we compared the estimation and inference from BAnOCC with that from CCLasso [9] , a frequentist LASSO-based method that chooses the shrinkage parameter using K-fold cross validation ( Fig 3 ) . BAnOCC had much lower false positive rates than CCLasso , resulting from the model’s ability to use the posterior distribution to account for estimate uncertainty while CCLasso , being LASSO-based , used a non-zero point estimate to determine significance of an effect . BAnOCC and CCLasso both estimate the log-basis correlation matrix accurately , and both are a substantial improvement on a naïve approach ( row 2 of Fig 3 ) . In particular , both BAnOCC and CCLasso have much lower false positive rates than Pearson correlation . Over all the null associations , Pearson correlation had a staggering false positive rate of 82%; CCLasso had almost 14% false positives as a result of many small but non-zero estimates; BAnOCC , because it uses the posterior credible intervals to evaluate uncertainty , had a false positive rate of about 3% . BAnOCC cannot estimate the log-basis correlations wjk to be exactly zero because of the continuous prior , but the null associations whose 95% credible intervals cover zero have very small estimates ( all are less than 0 . 15 , 75% are less than 0 . 05 ) . The association between features 1 and 5 in the “reversed spike” dataset was difficult for both BAnOCC and CCLasso . Both gave a small , negative estimate ( -0 . 001 for BAnOCC and -0 . 113 for CCLasso ) . BAnOCC displays a slight bias toward positive correlations instead of the moderate negative correlation that was present in the underlying unobserved abundances , as shown by several false positive associations in this dataset . This behavior is common among many methods , including SparCC and SPIEC-EASI ( S5 Fig ) . It results from the fact that when a negative-dominant structure is present , positive correlations become much more likely to be real than negative ones , an interesting observation to consider when interpreting real-world results from any of these methods . BAnOCC and CCLasso agree well with the true magnitude and direction of the non-zero associations that both methods conclude are significant . For these associations , the relative difference with the true value is less than 15% for both methods . When the associations were rejected , the 95% credible interval from BAnOCC covered the true value , indicating its utility for evaluating the uncertainty of the estimate . The false negative rates were 25% for BAnOCC and 0% for CCLasso , a direct result of the higher tolerance for false positives CCLasso exhibits . In practice , this has the expected effect of dramatically lowering BAnOCC’s false positive rate in recovering true correlations from compositional data . We compared BAnOCC’s performance as measured by type I and type II error rates to a range of previous methods ( Fig 4 ) : simplicial variation [23] , SparCC [8] , CCLasso [9] , SPIEC-EASI [11] , ReBoot [7] , and Spearman correlation ( directly on the composition as a negative control ) . Of the two frequentist LASSO-based methods ( CCLasso and REBACCA [10] ) , CCLasso alone had an R package interface; because they employ highly similar approaches , they should yield similar results . For a positive control , we also applied Spearman correlation to the unconstrained ( and usually unobserved ) counts ( Table 1 and S3 Text ) . Overall , BAnOCC controlled the type I error rate for all correlation strengths ( Fig 4A ) while maintaining comparable power compared with other recent methods ( Fig 4B ) . These results held true in a more even community with larger features , in which BAnOCC was the sole method to fully control the type I error rate ( S8 Fig ) . As the number of samples increased , all methods increased in power ( S9 Fig ) , while the type I error rates remained fairly constant ( S10 Fig ) . Only BAnOCC and SparCC controlled type I error while maintaining high power for all correlation strengths ( see also AUC boxplots in S6 Fig ) . Both behaved similarly to Spearman correlation applied to the unconstrained abundances , which represents the best possible performance ( as it uses the unconstrained data rather than the composition—this is impossible in practice , when only the composition is available ) . SparCC’s type I error rate was slightly inflated in a larger dataset with more features , while BAnOCC continued to control the type I error rate at the nominal level ( S8 Fig ) . As other authors have noted , SparCC does not guarantee that its log-basis correlation estimate has bounded elements nor that it is positive definite [9] . By contrast , BAnOCC not only estimates a positive definite correlation matrix with bounded elements , but also can infer network edges based on the precision matrix as well . Several methods proved to control the type I error rate poorly: Spearman correlation exemplifies this as a negative control , but simplicial variation , SPIEC-EASI using GLASSO and to a lesser extent CCLasso were comparable . ReBoot , by design , attenuates the type I error rate of Spearman correlation , but does not control it perfectly . The high type I error rates are also somewhat expected in simplicial variation , but SPIEC-EASI using GLASSO may not be performing as expected , especially since in contrast the Meinshausen-Bühlmann neighborhood selection method did control type I error . This may also possibly be because the neighborhood selection infers each element of the matrix one at a time , while GLASSO infers the matrix all at once; this makes the GLASSO optimization a more difficult problem . Feature 5 in the template dataset has a large mean and variance , while feature 3 has a small mean and variance . This results in a strong negative spurious correlation in the composition , which gives rise to interesting behavior of essentially all methods when detecting this association . When the true association is negative , many compositionally-appropriate methods such as BAnOCC , SparCC , and SPIEC-EASI ( MB ) do poorly at detecting the true correlation ( Fig 4B ) because the negative correlation is difficult to attribute to the unobserved counts rather than spurious correlation . Conversely , more naïve methods such as simplicial variation and Spearman correlation do very well at detecting a weak negative correlation between these two features because this becomes a strong negative correlation in the composition . This simulated example thus provides some insight into the form of sensitivity / specificity tradeoff that applies in the constrained , information-loss setting of identifying true correlations from compositions .
As an example application , we inferred a correlation network among microbial taxa profiled using ecological data from the Human Microbiome Project [22] ( Fig 5 ) . Microbial community sequencing generates compositions by assigning sequencing reads to microorganisms; since nucleotide sequencing depth is arbitrary , the resulting counts are not informative regarding the unobserved and unconstrained counts and are often normalized to relative abundances . Co-variation patterns in such data are of interest because they suggest ecological interactions , such as mutualism ( positive correlation ) or predation ( negative correlation ) [7] . The microbial taxonomic relative abundance data used here consisted of 523 microbial features measured across 700 total samples using MetaPhlAn2 v2 . 0_beta1 [24] in July of 2014 ( available in S3 Data ) , further excluding from all networks markers removed in the subsequent version’s database ( v2 . 0_beta2 ) . These samples were in turn drawn from 127 individuals at six distinct body sites . Microbial ecology differs at each body site [22] , providing examples for BAnOCC analysis that ranged from diverse , relatively even communities ( such as stool ) to less diverse , highly skewed ecologies ( such as the vaginal posterior fornix ) . For each of three representative body sites ( stool , posterior fornix , and buccal mucosa ) , we selected the first time point from each subject , collapsed taxonomic information to the genus level , and then removed features with relative abundance less than 0 . 0001 in at least 50% of samples . With too few features , little to nothing can be concluded about the true correlations; so if fewer than 10 features remained we lowered the prevalence cutoff until 10 features were retained . The hyperparameters for the gamma prior on λ were a = 0 . 5 and b = 5 for all body sites , ensuring that we gave substantial weight to sparser precision matrices . For all body sites , we used the prior variability of the log-basis means L = 30I; each body site , however , had a different nj so that the distribution of the sums of medians were similar across different body sites ( see S11–S14 Figs ) . We further compared BAnOCC’s inferred network using the log-basis correlation matrix with that from CCLasso , and BAnOCC’s inferred network using the log-basis precision matrix with SPIEC-EASI . There is broad agreement between the methods as to which edges are significant , with very few edges discrepant between the methods ( S15 Fig ) . In stool , BAnOCC inferred several positive associations between genera within the family Bacteroidales , in particular Bacteroides , Odoribacter , Parabacteroides and Alistipes ( Fig 5A ) . Until recently , these genera were classified as part of the same genus [25] . This supports the common observation that closely ( but not too closely ) related taxa tend to have positive ecological associations [26] . Additionally , positive associations in the buccal mucosa ( Fig 5B ) connect taxa that are known to physically co-aggregate; in particular , Fusobacterium interactions with species from the Porphyromonas and Capnocytophaga genera ( among others ) are crucial in biofilm formation [27] and have been previously recovered from 16S-based ecological analyses [7] . Lastly , we can see the well-documented negative association between the Lactobacillus genus in the posterior fornix with several genera associated with dysbiosis such as Gardnerella and Prevotella [28] ( Fig 5C ) . Two interactions newly suggested by this analysis involved the Proteobacteria across multiple body sites , and specifically in stool and the oral cavity ( buccal mucosa ) . The genera Escherichia and Haemophilus represent the two major proteobacterial residents in these habitats , respectively , and both were involved in predominantly negative interactions with more typical , abundant members of these communities ( e . g . Faecalibacterium and Eubacterium in the gut , Leptotrichia or Corynebacterium in the mouth ) . These clades are highly phylogenetically diverged and tend to carry larger , more generalize genomes and pan-genomes [29 , 30]; this suggests that they will overgrow in these habitats only in unusual situations , exemplified by E . coli’s abundance in the gut primarily during inflammation [31] . Further details may be provided by future analyses using BAnOCC or related methods on species or strain-level ecological profiling .
Here , we describe BAnOCC , a Bayesian method for inferring the log-basis correlation structure from compositional data . Assuming a log-normal distribution on the unobserved and unconstrained counts , the model estimates the log-basis correlations using a sparsity-inducing shrinkage prior on the log-basis precision matrix . It is part of a family of several recently proposed LASSO-based methods [9–11] which provide a more rigorous approach to correcting for compositional effects than earlier methods [7 , 8] . Unlike the other LASSO-based correlation-inference methods that summarize pairwise associations using a single point estimate , BAnOCC yields uncertainty estimates of the precision , covariance , and correlation parameters . Simulation results show that BAnOCC performs as well as or better than existing methods in controlling type I error while maintaining power for network edge detection from compositional data . Finally , we applied the method to assess microbial relationships in the human microbiome , confirming established interactions and suggesting novel ones for future validation . Analysis using a Taylor series approximation provided one of the first characterizations of properties that make true correlations “difficult” to recover from compositions , or conversely “easy” to miss as false negatives . In particular , this depends not only on the more intuitive number and evenness of feature means , but also on the distribution of their variance . This allowed us to simulate designedly difficult test cases for BAnOCC and a variety of published methods , in contrast to previous simulation studies that relied primarily on relatively simple synthetic data [7–10] . In most studies , spurious correlation is noted to be commonly present and of varying magnitudes and directions [11] . However , the possible sensitivity of methods to the type of spurious correlation encountered has not been explored and is an important contribution to the characterization of existing and future methods . We anticipate several computational and statistical refinements that may further improve BAnOCC’s performance . While BAnOCC uses 95% credible intervals for inference , these can be overly conservative [32] . Alternative thresholding methods may improve on this , such as the scaled neighborhood criterion [32] or the partial-correlation based approach of [33] and [13] . A discrete-continuous mixture prior such as the G -Wishart prior [34] or the covariance selection prior [35] on the log-basis correlation matrix would further allow the posterior probability that wjk = 0 to be nonzero , and this quantity could be used as a threshold . For applications specifically on count data , such as microbial compositions , the data could be modeled more accurately by adding a hierarchical layer . This would generate measurement counts conditional on the unobserved and unconstrained counts , making the observed compositions a function of normalized measurement counts . The degree of zero-inflation observed in ecological data could also be modeled directly using a hurdle or mixture model , or a multinomial distribution for the measurement counts . This would provide a particularly targeted approach for microbial ecology , in which more detailed data ( at the species or strain level [24] ) could be further incorporated . We thus hope to refine both the accuracy of compositional correlation inference and the applications to microbial community data in future studies .
|
Data from many fields are available primarily in the form of proportions , also referred to as compositions , which impose mathematical constraints on identifying interactions among components in the underlying systems . In particular , correlations cannot be calculated directly from proportions or from count data that give rise to them . Methods that work around this difficulty generally do so by imposing strong assumptions about the distribution of underlying data or associated correlations , and these in turn often prevent quantifying uncertainty in the resulting estimates of correlation . We developed a statistical model ( BAnOCC: Bayesian Analysis of Compositional Covariance ) that both estimates correlations between counts or proportions and provides a posterior distribution for each correlation that quantifies how uncertain the estimate is . BAnOCC does well at controlling the number of false positives in simulated data and can be practically applied to a wide range of proportional data types .
|
[
"Abstract",
"Introduction",
"Methods",
"BAnOCC:",
"Bayesian",
"analysis",
"of",
"compositional",
"covariance",
"Choosing",
"hyperparameters",
"Software",
"Results",
"Simulation",
"studies",
"A",
"microbial",
"interaction",
"network",
"from",
"the",
"Human",
"Microbiome",
"Project",
"Discussion"
] |
[
"ecology",
"and",
"environmental",
"sciences",
"microbiome",
"microbiology",
"random",
"variables",
"covariance",
"simulation",
"and",
"modeling",
"mathematics",
"theoretical",
"ecology",
"microbial",
"genomics",
"research",
"and",
"analysis",
"methods",
"medical",
"microbiology",
"mathematical",
"and",
"statistical",
"techniques",
"bayesian",
"method",
"research",
"assessment",
"microbial",
"ecology",
"probability",
"theory",
"community",
"ecology",
"ecology",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"genomics",
"research",
"errors"
] |
2017
|
A Bayesian method for detecting pairwise associations in compositional data
|
Intestinal nematodes suppress immune responses in the context of allergy , gut inflammation , secondary infection and vaccination . Several mechanisms have been proposed for this suppression including alterations in Th2 cell differentiation and increased Treg cell suppressive function . In this study , we show that chronic nematode infection leads to reduced peripheral responses to vaccination because of a generalized reduction in the available responsive lymphocyte pool . We found that superficial skin-draining lymph nodes ( LNs ) in mice that are chronically infected with the intestinal nematode Heligmosomides polygyrus , do not reach the same cellularity as worm-free mice upon subsequent BCG infection in the skin . B cells and T cells , all declined in skin-draining LN of H . polygyrus-infected mice , resulting in LNs atrophy and altered lymphocyte composition . Importantly , anti-helminthic treatment improved lymphocyte numbers in skin-draining LN , indicating that time after de-worming is critical to regain full-scale LN cellularity . De-worming , and time for the skin LN to recover cellularity , also mended responses to Bacille Calmette-Guerin ( BCG ) in the LN draining the footpad injection site . Thus , our findings show that chronic nematode infection leads to a paucity of lymphocytes in peripheral lymph nodes , which acts to reduce the efficacy of immune responses at these sites .
Infections with intestinal nematodes often become chronic in mammals due to both the longevity of worms and continuous reinfection . Worm infections , once chronically established , typically cause little overt pathology but may have implications on growth development , nutritional status and the ability to mount secondary immune responses . Suppressed lymphocyte responses in individuals with worm infections were described around 40 years ago and have since been confirmed in both infectious and autoimmune disease animal models [1–4] . Importantly , de-worming people prior to vaccination may be a strategy to improve vaccine and therapeutic efficacy in populations with high worm-burden [5–7] . Worms have been well-characterised to induce potent Th2 cell responses , typically characterised by the secretion of cytokines such as IL-4 and IL-13 , which promote nematode clearance in a number of ways [8] . Viruses and bacterial species including mycobacteria induce Th1 cell responses , which help to eliminate these intracellular pathogens . These two arms of the CD4 T cell response are antagonistic to one another , and as such , one presumption has been that Th2-biased chronic worm infections prevent the efficient development of robust Th1 immunity , required for example , in mycobacteria vaccine responses [9] . Furthermore , helminths are also potent inducers of regulatory T cell ( Treg ) responses , which have a generalized inhibitory effect on immune function [10] . Downregulation of inflammatory immune responses is essential to minimize pathology in the host , but benefits the survival and chronic establishment of the worm . Despite the regulatory responses evoked by intestinal worms , inflammatory reactions are not completely abolished and a persistent worm infection will provide a continuous supply of antigens that maintain stimulation to the immune system [11] . Many of the stereotypic immune characteristics of a helminth infection including Th2 cells , hallmarked by their production of IL-4 , IL-5 and IL-13 , regulatory Foxp3+ T cells producing TGF-β and IL-10 , goblet hyperplasia , mucus production , mast cell activation and differentiation of M2 macrophages [12] occur in proximity to the worm infection . Thus , the immune responses that directly could counteract responses to vaccination and / or co-infections are primarily compartmentalized to the gut and associated lymphoid tissue . However , human studies as well as our own experimental data show that intestinal nematodes have inhibitory effects on immune responses to BCG immunization given in the skin [2 , 7] . Thus , it is evident that an intestinal worm also has widespread systemic effects . It is not clear how these distal inhibitory effects are mediated . Herein we comprehensively analysed how distal skin-draining LN are affected by chronic intestinal worm infection . We found no evidence that Th2 or regulatory T cells disseminated throughout the organism , such that they would directly inhibit immune responses to BCG infection in the skin . Instead , our data support the notion that intestinal worms cause atrophy of skin-draining LN . Skin-draining LN , as they decrease in cellularity , also change in composition having a lower T cell/B cell ratio . Upon secondary immunization or infection , the LN draining the second inoculation site is smaller in worm-infected compared to worm-free mice . These LN , being smaller , have a reduced capacity to recruit and retain cells , resulting in a diminished expansion in response to secondary immunization/infection . Administration of recombinant IL-7 ( rIL-7 ) , which increases the survival of lymphocytes and supports an expansion of the total lymphocyte pool [13] , allowed worm-infected mice to maintain a larger cell number in skin-draining LN without additional expansion of mesenteric LN ( mLN ) . De-worming of mice led to a recovery of distal LN cellularity , as the mLN contracted . Our findings suggest that the distal immune suppressive effects of intestinal nematodes are not due to specific effects of regulatory and Th2 responses evoked by worms , but rather a broader reconfiguration of the lymphoid system . Our data also explain why BCG vaccination may be less effective in individuals with intestinal helminths and suggest that de-worming may recover cellular responses . However , some time may be needed before effects of de-worming can be beneficial .
We have previously reported that mice with chronic H . polygyrus infection have reduced cellular responses when subsequently infected with BCG [2] . To see how the BCG response developed in worm-infected and worm-free mice respectively , we examined the cellular expansion of the reactive popliteal LN ( pLN ) following infection with BCG in the footpad . In line with our previous observations , LN expansion in response to BCG was reduced in mice with chronic H . polygyrus co-infection and did not reach the cellularity of worm-free mice during 12 days ( Fig 1A ) . In response to BCG footpad infection , fewer dendritic cells ( DCs ) migrated from the footpad to the draining popliteal LN ( pLN ) in worm-infected mice compared to worm-free mice . Likewise , fewer mycobacteria specific CD4 T cells ( P25-TCRTg cells ) were found in the pLN of worm-infected compared to worm-free mice after BCG infection . Although , there was a difference in the total number cells in response to BCG infection neither the migratory DCs nor the mycobacteria-specific P25-TCRTg cells appear to be specifically targeted . The frequencies of afore mentioned cells were in fact similar in pLN of worm-infected and worm-free mice infected with BCG in the footpad . When further calculated , it was evident that both the number of DCs migrating from the skin to the draining LN 48–72 hours after BCG infection ( Fig 1B ) and the number of P25-TCRTg T cells measured 6 days after infection ( Fig 1C ) are directly proportion to the total cellularity of the lymph node . This correlation was independent of worm infection . The induction of regulatory T cells ( Treg ) has been proposed as a major mechanism by which worms down-modulate host immunity [4] . To assess whether H . polygyrus induced widespread systemic immune suppression , we evaluated Treg and cytokine expression in skin-draining LN . As expected , an increase in regulatory Foxp3+ CD4+ T cells was evident in the mLN following H . polygyrus infection ( Fig 2A and 2B ) . This increase did not extend to skin-draining LN where the frequency of Foxp3-expressing CD4+ cells was unchanged and cell numbers actually decreased ( Fig 2C and 2D ) . In line with this , RNA sequencing of distal LN in control or chronic H . polygyrus infection did not reveal any specific up- or down-regulation of Th2- or Treg-associated genes ( Fig 2E ) . Further , targeted RT-PCR analysis of relative gene expression associated with worm-induced immune responses did not show up-regulation of gene expression associated with Th2 ( GATA-3 , IL-4 ) or Treg cells ( IL-10 , TGF-β ) in skin-draining LN of H . polygyrus-infected mice ( Fig 2F and 2G ) . Others have indicated that regulatory and Th2 responses induced by the worm are evident in the spleen [14] . We do not exclude that this may occur , but our data does not support heightened Treg or Th2 responses in the spleen . We have shown previously [2] that frequencies and numbers of CD4+Foxp3+ T cells are similar in the spleens of worm-infected and worm-free mice . In line with this , we here show that splenic expressions of selected Th2 and Treg genes were similar H . polygyrus infected and control mice ( Fig 2H and 2I ) . We then sought to understand if other underlying changes in the superficial LN could explain the diminished responses to secondary infection/vaccination in the skin . The RNA-seq analysis pointed to a reduced lymphocyte cellularity and an increased expression of genes associated with matrix regulation in inguinal LN ( iLN ) of worm-infected mice ( Fig 2E ) . Consistent with this , we found the cellularity of skin-draining LN was less in mice with chronic H . polygyrus infection ( Fig 3A–3C ) . Both weight and cellularity were gradually reduced in iLN from H . polygyrus-infected mice ( Fig 3D and 3E ) , becoming significant three weeks after infection , partially in line with previous work [15] . At the time when the H . polygyrus infection was considered chronic ( 28 days post infection ) [12 , 16] , LN cellularity did not change further . That said , the atrophy of skin-draining LN was then preserved and still evident three months after H . polygyrus infection ( S1A Fig ) . The effect of H . polygyrus infection on skin-draining LN was partially dose-dependent , a lesser effect was observed following infection with 50 L3 while giving 100 L3 almost had the same effect as 200 L3 ( S1B Fig ) . Smaller skin-draining LN observed in H . polygyrus-infected mice could not be explained by impaired growth development since infected mice gained weight normally ( S1C Fig ) and consumed similar amounts of food as non-infected mice ( S1D Fig ) . Further , skin LN atrophy was not a consequence of altered immune cell distribution in growing mice since smaller skin-draining LN were also evident in adult ( 9 weeks old at time of infection ) mice infected with H . polygyrus ( S1E Fig ) . The overall loss of cellularity in skin-draining LN of H . polygyrus-infected mice was primarily in lymphocytes ( Fig 3F ) . Stromal and myeloid cells were not significantly affected , though the trend was toward a decrease also in these populations ( Fig 3F ) . Histopathological evaluation of superficial LN did not point to any major structural changes in H . polygyrus infected mice ( Fig 3G ) . On the other hand , the effect on lymphocyte subsets was not equally distributed . More T cells than B cells were lost and viewed as frequencies CD8+ T cells were lost over CD4+ T cells , causing an alteration in T cell/B cell as well as CD4+/CD8+ T cell ratio in skin-draining LN of H . polygyrus-infected compared to worm-free mice ( Fig 3H and 3I ) . Analysis of CD44 , CD62L and CD103 on T cells from skin-draining LN show that H . polygyrus infection caused a general reduction in T cells , although the loss of naïve T cells ( CD4+CD62Lhi CD44int/low and CD8+ CD62Lhi CD44int/low as well as CD8+ CD103int T cells ) was most evident ( Fig 3J–3L , S1F–S1O Fig ) . This suggests that naïve lymphocytes primarily are lost from skin-draining lymph nodes in chronic worm infection . Considering that naïve lymphocytes were more affected than other subsets , we measured the expression the integrin GlyCAM-1 and ICAM-1 as well as the chemokines CCL19 and CCL21 , and the cytokine IL-7 , since all are important in recruitment and maintenance of lymphocytes in the skin-draining LN [17] . However , we found no change in the relative expression of these molecules in iLN between H . polygyrus-infected and uninfected mice ( Fig 3M–3O ) . The maintenance of LN cellularity is mainly determined by input and output of circulating lymphocytes between lymphoid organs and blood [18] . Thus , smaller skin-draining LNs may reflect changes in the concentration of circulating blood lymphocytes . To address this we followed blood lymphocyte counts following H . polygyrus infection . In support that the smaller LN are a consequence of decreased lymphocyte trafficking , we found that the levels of circulating lymphocytes drop in mice following H . polygyrus infection ( Fig 4A ) . While the decline of cellularity in skin-draining LN appeared continuous , assessment of blood lymphocytes indicated that the reduction might occur in waves . Blood lymphopenia during H . polygyrus infection was most evident in the first days after infection and 2–3 weeks after infection with a period of normal values in between ( Fig 4A ) . The drops in blood lymphocyte levels coincided relatively well with described waves of innate and Th2 immune responses towards the worm in C57BL/6 mice and accompanying enlargement of the reactive mLN [16] . Upon transfer of lymphocytes from an uninfected mouse , fewer donor cells were found in the circulation of the worm-infected compared to the worm-free recipient mice 4 hours after transfer ( Fig 4B ) . This indicated that lymphocytes might be more rapidly removed from the circulation in mice with chronic H . polygyrus infection . The superficial LN atrophy and lymphopenia did not appear to be coupled to thymic involution observed in other infections [19–22] , since the weight and cellularity of the thymus as well as Ki67 staining of thymocytes were not changed by chronic H . polygyrus infection ( Fig 4C–4E ) . The mLN draining the intestine displayed a 5-fold increase in cellularity in mice chronically infected with H . polygyrus compared to uninfected mice ( Fig 5A ) . The cellularity of the spleen was on the other hand not significantly changed ( S2A Fig ) . While the majority of CD4+ T cells in mLN of infected mice had a naïve phenotype ( CD62LhiCD44int/low ) , as expected , more effector ( CD62LlowCD44hi ) cells were also found in worm-infected animals compared to worm-free mice ( Fig 5B–5E ) . To test if naïve cells were “trapped” in the mLN of infected mice , we transferred total labelled lymphocytes from uninfected animals into recipients with or without chronic H . polygyrus infection . We tracked these transferred cells and found a higher number in mLN and a lower number in skin-draining LN of H . polygyrus-infected compared to control mice 4 hours after transfer ( Fig 5F and 5G ) . However , when viewed as frequency , it was evident that the relative numbers and the cell subsets entering the superficial LN were similar and proportional to the total cellularity of the various LN ( Fig 5H and 5I , S2B–S2F Fig ) . Similar to that observed in skin-draining LN , the frequency of transferred T cells found in mLN were not altered by the worm infection , and only reflected the proportion of these subsets in the transferred population ( S2G , S2H , S2J and S2K Fig ) , while the frequency of transferred B cells were slightly higher in mLN ( S2I Fig ) . Tracking of labelled cells 12 hours after transfer showed a similar distribution in skin-draining LN and mLN as during 4 hours ( S2L–S2O Fig ) . To test how cells with antigen specificity unrelated to the worm would distribute in H . polygyrus-infected animals over time , we transferred mycobacteria Ag85B-specific P25-TCRTg cells to mice prior to worm infection . As was the case for the transfer of total lymphocytes ( Fig 5F and 5G ) , P25-TCRTg cells were maintained in higher numbers in mLN and in lower numbers in skin-draining LN of mice with chronic worm-infection compared to uninfected mice ( Fig 5J and 5K ) . Our data show that skin-draining LN become atrophic as the worm infection progresses and the mLN expand . This could be an effect of fewer circulating cells reaching the skin-draining LN , possibly being sequestered to the enlarged mLN and the intestine . In an attempt to assess the impact of circulating lymphocytes on LN cellularity during the span of experimental H . polygyrus infection , we systemically blocked lymphocyte egress throughout the course of the infection using the S1P1 agonist FTY720 . This treatment removes lymphocytes from the circulation and prevent cells from leaving the bone marrow , the thymus and LN [23 , 24] . Long-term treatment with FTY720 is quite safe , the toxic effects , apart from blocking lymphocyte egress are mainly on cancerous cells leaving normal cells intact . As expected , much fewer lymphocytes were detectable in blood after FTY720 treatment ( 5–6 times less compared to untreated ) . The FTY720 treatment alone , given over 4 weeks , caused a generalized reduction in cellularity in both mLN and iLN ( Fig 5L and 5M ) . After 4 weeks of FTY720 treatment the skin-LN cellularity was about 50% of that in untreated mice regardless of H . polygyrus infection or not ( Fig 5L ) . The mLN in H . polygyrus-infected and FTY720-treated mice were expanded ( Fig 5M ) , albeit much less compared to the untreated H . polygyrus infected mice ( Fig 5M ) . As anticipated FTY720 dramatically affected naïve cells in particular CD4 T cells ( Fig 5N ) . Weight gain and worm-load was similar in FTY720 treated and untreated animals . We believe these results reflect the requirement of a constant supply of circulating naïve lymphocytes for maintenance and expansion of LN , in presence or absence of worm infection . To test if the loss of cellularity in superficial LN could be compensated for in H . polygyrus-infected mice , we transferred large numbers of lymphocytes at regular intervals following infection . This procedure did not alter cellularity in superficial LN from worm-infected or uninfected animals ( Fig 6A ) . This suggests that in the case of chronic infection , lymphoid homeostasis is altered to accommodate for the expansion of the mLN , at the expense of the cellularity of peripheral LN , and that this peripheral deficit cannot be rescued simply by providing more cells . We then questioned whether we could restore peripheral LN cellularity by supplying factors involved in lymphoid homeostasis . To this end , we administered recombinant IL-7 to mice throughout the course of H . polygyrus infection . The idea behind the IL-7 supplementation was to allow a larger lymphocyte pool . IL-7 is essential for survival of lymphocytes and limitations in IL-7 could prevent the lymphocyte pool from expanding and be a regulating factor in maintenance of LN cellularity [13] . rIL-7 treatment led to an increase in cellularity and recovery of T and B the iLN ( Fig 6B and 6C , S3A–S3H Fig ) while the mLN did not expand further in H . polygyrus-infected mice ( Fig 6D ) , supporting the notion that the size of the lymphocyte pool may be a limiting factor in determining LN size both during steady state and infection . Killing of parasitic worms have been proposed as a strategy to improve vaccine efficacy [5 , 25] . To test if removal of worms would lead to recovery of LN cellularity we treated mice chronically infected with H . polygyrus with three doses of pyrantel pamoate over 5 days . LN cellularity was assessed 10- and 21- days after therapy . De-worming removed worms within 2–3 days , as determined by the lack of eggs in faecal samples . The enlargement of the mLN persisted for weeks ( Fig 6E and 6F ) . This we propose can have implications on skin-draining LN , which remained smaller , compared to control mice and similar in cellularity and composition to mice infected with worms 10 days after de-worming ( Fig 6G and S3I–S3P Fig ) . Three weeks after anti-helminthic treatment , the cellularity of iLN had increased and the number of T and B cells were similar to that of control mice ( Fig 6H , S3Q–S3X Fig ) . We then addressed if de-worming and time given for the LN to regain cellularity would correct the impaired responses to BCG previously shown in Fig 1 and [2] . Indeed , we found that responses to BCG were returning and were similar in de-wormed and worm free animals ( Fig 6I–6L ) . Indicating that de-worming , given time , positively affect the ability to mount antigen specific response to immunization in the skin .
Chronic helminth infections are implicated in impaired responses to vaccination and control of secondary infections such as tuberculosis [9 , 26 , 27] . We have previously shown that mice with chronic H . polygyrus infection have muted responses to BCG and Leishmania infection [2] . Th2 and regulatory T-cell responses are typically depicted as the culprits causing the worm-mediated suppression of Th1-controlled infections [27–31] . However , we find no evidence for disseminated up-regulation of Th2 or regulatory cytokines in skin LN . Nor is there a widespread increase in regulatory T -cells in our model [2] ( Fig 2F and 2G ) . Instead , our data suggests that distal immune suppression in worm-infected individuals may be a consequence of redistribution and competition for the available lymphocyte pool . Our interpretation is supported by a recent study showing that naïve lymphocytes accumulate in the mLN and are lost from peripheral LN in worm-infected mice [15] . Similar to our observations using BCG , these mice displayed muted responses and impaired control of influenza infection . Impaired immune responses to skin-BCG infection can thus be an effect of reduced cellularity of the BCG draining node . There is a positive correlation between LN size and response to BCG independent of worm infection . Mice with intestinal worms , which have smaller skin-draining LN , accordingly displayed a reduced response to BCG injection in the footpad compared to worm-free mice ( Fig 1 ) . The loss of cellularity in skin-draining LN was partially dose-dependent and less obvious when a low-dose worm infection was used ( S1B Fig ) , supporting the notion that negative effects of worms are most evident when the worm-burden is high . Other than size , we found no gross pathological alterations evident macroscopically or by histology in the superficial LN of worm-infected mice . That said , lymphocyte numbers were reduced in the skin-draining LN of worm-infected mice , where furthermore , an increase in B-cell/T-cell and CD4+/CD8+ T-cell ratios were seen . The latter observation likely reflects the turnover of the respective cells subsets within a LN , B cells having a more mature phenotype and thus slower turnover than T cells . Further , naïve cells were lost over memory/effector cells , similar to findings by King et al . [15] . How these changes in lymphocytes composition affect subsequent immune responses in detail remain to be determined . Thymic involution and decreased output of T cells has been reported to occur in infectious as well as inflammatory diseases [21] . The lymphocyte levels fluctuated during H . polygyrus infection . However , our results suggest that the thymus is normal in mice with chronic H . polygyrus infection , with similar weight , cellularity and proliferative capacity of thymocytes as that of uninfected animals ( Fig 4 ) . Thus , the reduced cellularity in superficial LN is not likely due to reduced thymic output . In addition , we did not find any evidence for increased cell death in skin-draining LN from worm-infected mice and mRNA expressions of caspase-3 were not different between the groups . The reactive mLN draining the small intestine was greatly enlarged in mice with chronic H . polygyrus infection ( Fig 5A ) , and remained so for as long as we followed the infection ( 3 months ) . A LN expands in several steps . Upon infection , the draining LN rapidly responds to accommodate an expansion in cellularity . More lymphocytes are allowed to enter the LN through the high endothelial venules and signals that allow lymphocytes to egress from the LN are down regulated resulting in a “LN shut down” [32] . This “trapping” of naïve lymphocytes results in a rapid increase in cellularity of the reactive LN [17 , 33] serving to increase the probability of naïve T-cell encounter with their cognate antigen on antigen presenting cells thereby facilitating the initiation of an adaptive immune response [33] . The LN then further expand as lymphocytes undergo clonal expansion . In a resolving infection , where the pathogen is cleared , the immune response contracts and the reactive LN returns to normal size . In a chronic intestinal nematode infection this does not occur . Rather , the infection persists and the mLN remain enlarged . The expansion of mLN and the accompanying reduction of skin-draining LN was dependent on the infectious dose ( S1B Fig ) . Thus , the immune response mounted , measured as total cellularity , match the worm load . Interestingly , following de-worming , the mLN remain enlarged for weeks . In viral infections , it has been suggested that antigens can persist for extended time even in the absences of the infection and maintain T cell responses [34] . However , other infections may require persistence of the infection to uphold T cell responses . CD4 T cells appear to require a continuous supply of protein/peptide to remain activated , and following removal of the antigen source the ability of DCs from draining LN to activate antigen specific responses are rapidly lost [35] . While persisting worm antigen and/or reminiscent inflammation may play a part in maintenance of mLN sizes following de-worming , we suggest that the enlarged LN size per se is as important . Size dependent distribution and re-distribution of lymphocytes amongst LN , alone , can explain why skin-draining LN loose cellularity and why it takes time to regain peripheral LN cellularity in previously worm-infected mice . Assuming that the total lymphocyte output and pool are relatively constant [36] and , as we show , that lymphocytes distribute amongst LN according to LN cellularity/size our data suggest that the mLN simply by being enlarged will over time cause skin-draining LN to lose cellularity . A larger node can retain more of the circulating lymphocytes . The physical properties that accompanies a larger LN size , such as longer cell retention time in the LN , will be important in maintaining higher cellularity when lymphocyte numbers are limited . Thus , unless more lymphocytes are produced , or survival increased , expansion of one LN will be at the expense of others . We consider thus , that distribution of circulating lymphocytes is foremost a function of LN size/cellularity . Accordingly , transfer of cells to H . polygyrus infected mice showed that lymphocytes distribute proportional to the size of skin-draining LN size just as they do in uninfected mice , with fewer transferred cells found in skin-draining LN of worm-infected mice as they are smaller and the enlarged mLN contained more transferred cells compared to uninfected mice ( Fig 5F–5K ) . Thus , the LN size per se contribute to regulate the number of cells in a LN during steady state as well as upon chronic infection . Since mLN cellularity decrease slowly following termination of infection , the capacity to retain circulating lymphocytes , while steadily decreasing , can remain for a prolonged period . On this line of thought , resolution of skin LN cellularity after a worm infection is bound to take time , even in the absence of persisting antigens . Mesenteric LN expansion in response to H . polygyrus infection is however constrained . We found that , while many more in absolute numbers , the frequencies of transferred cells were actually less in mLN of H . polygyrus infected mice . This may reflect that counter regulatory mechanisms are in place preventing the mLN to expand overly much . Indeed , we found , in line with previous reports , that the relative expression of both CCL21 and CCL19 mRNAs were down regulated in the reactive mLN and about 50% of that in uninfected mice , [37] . Moreover , treatment with rIL-7 , while expanding the superficial LN in worm-infected mice , did not further expand the already enlarged mLN ( Fig 6B–6D ) . The upkeep of LN is dependent on a continuous supply of new lymphocytes . Without this refill , the number of lymphocytes decline and LN lose size and cellularity , as observed following FTY720 treatment . Limitations in the output of lymphocytes and/or size of the lymphocyte pool may thus explain the skin LN atrophy observed in H . polygyrus infected mice . However , transfer of large number of lymphocytes , to compensate for the loss of cellularity in superficial LN of H . polygyrus infected mice , did not alter LN cellularity . This is in line with that lymphocytes , at least under steady state conditions , can only exist in predestined quantities and that the total number of lymphocytes within the circulation and secondary lymphoid tissue is restricted [36] . IL-7 is essential in lymphocyte homeostasis and is a limiting factor in determining the size of the lymphocyte pool [13] . The circulating levels of IL-7 were very low ( 0–10 pg ) and mRNA expression levels in iLN similar in worm-infected and worm-free mice ( Fig 3N ) . Yet , administration of exogenous IL-7 would allow an expansion of the lymphocyte pool and resulted in our system in larger inguinal LN , with restored naïve T cell and B cell numbers , in worm-infected mice ( S3A–S3H Fig ) . In our system , the host immune response to the infection did not compensate for the reduction of lymphocytes in superficial skin LN . As such , the expansion of mLN in the gut occur at the expense of peripheral LN cellularity , with a negative correlation between iLN and mLN sizes following H . polygyrus infection ( Fig 7A ) . Thus , the muted immune response to BCG infection in skin is likely explained by the overall reduction in lymphocyte numbers in peripheral skin LN of worm-infected individuals . A model in which a redistribution of the circulating lymphocytes amongst LN occurs following H . polygyrus infection and is maintained in the chronic phase by the sizes of LN can be envisage ( Fig 7B ) . We can assume that there is a cost benefit to limiting the size of the lymphocyte pool , saving the host’s energy as well as decreasing the risks of immune-associated pathologies . However , limiting lymphocyte numbers may compromise the ability of the host to mount multiple immune reactions . In chronic intestinal worm infection , the atrophy in superficial LN can result in a reduced capacity to initiate an immune response elsewhere resulting in a diminished immune response to a secondary infection in the periphery . This competition for cells provide an explanation to why host immunity to co-infections and vaccination can be dampened in animals , including humans , with chronic intestinal worms , which does not involve Th2 or regulatory cells . The competition for lymphocytes in our model does not exclude that circulating cells also redistribute to Peyer’s patches ( PP ) or the small intestine ( SI ) . Such involvement would not change our interpretation , but rather more broadly implicate the gut associated lymphoid tissue in the redistribution of lymphocytes during chronic infection with intestinal worms . However , the mLN are most likely the main receivers of redistributes cells , since the number of lymphocytes in PP and SI are fewer compared to the mLN [38] , thus their contribution less according to our model . Further local immune dampening effects by the worms may also have a more direct effect in the SI preventing PP from expanding overly much [39] . De-worming may be a strategy to boost responsiveness to vaccines in areas where high worm burdens prevails . In support of this , worm-infected individuals treated with albendazol were found to have better immune responses to BCG and malaria parasites compared to those left untreated [7 , 25] . Our data show that de-worming can restore LN cellularity and responses to a BCG injection given in the skin . Importantly our data show that time is needed for the peripheral LN to recover cellularity . Indicating that time between de-worming and vaccination may be an important consideration for maximizing the outcome of the subsequent vaccination .
C57BL/6 congenic CD45 . 1 ( Ly5 . 1 ) , C57BL/6-GFP , P25-TCRTg RAG-1-/- [40] x RAG-1-/- ECFP or EGFP ( originally provided by Dr . R . Germain , NIAID , USA ) mice were bred and maintained under specific-pathogen-free conditions at MTC or KM-Wallenberg facilities , Karolinska Institutet ( KI ) , Sweden . Wild type C57BL/6 mice were either bred at MTC/KI or purchased from Janvier Labs ( Renne , France ) . Both female and male mice were used . Mice were age and sex matched . All infections were performed in wild type ( C57BL/6 or congenic Ly5 . 1/CD45 . 1 ) mice . If not otherwise mentioned mice were infected by oral gavage with 200 H . polygyrus L3 larvae , obtained as described previously [41 , 42] at 4–5 weeks of age . The worm-infections were considered chronic after 28 days . At the end of each experiment , the worm burden was estimated by counting viable worms that had migrated out of the opened intestine through a fine net into a tube containing RPMI-1640 at 37°C within 3–4 hours . Mycobacterium bovis Bacillus Calmette-Guérin ( BCG ) strain SSI 1331 ( Statens Serum Institut , Denmark ) was expanded in 7H9 medium and inoculated at 1×106 colony forming units ( CFU ) in the footpad as described elsewhere [43] . Superficial skin-draining ( popliteal , pLN; inguinal , iLN; and axillary , aLN ) and total mesenteric LN ( mLN ) were collected in PBS at various time points following H . polygyrus and BCG infections . Lymph node ( LN ) weight was determined using an analytical balance ( Denver Instruments SI114 ) . Cellularity was estimated in single cells suspension of LN , prepared by crushing the LN with a pestle or to determine stromal cells following tissue digestion , as described elsewhere [44] . Lymphocytes were counted using trypan blue exclusion either by microscopy using a haemocytometer , by an automated cell counter cell ( Countess II , Life Technologies ) or by FACS using beads ( Countbright , Absolute bright count , Thermo Scientific ) . Since single cells suspension of LN are >95% lymphocytes ( S4 Fig ) and differences between counting methods was considered not to significantly impact the results and was not corrected for . For estimation of cells in blood , 20μl venous blood was collected from the tail into EDTA treated tubes at different times points following H . polygyrus infection . Cell counts were determined using an automated haematology analyser ( Mindray BC-2800Vet ) . Single-cell suspensions from tissues were incubated with 0 . 5 mg/ml anti-mouse FcγIII/II receptor ( 2 . 4G2 ) ( BD Biosciences ) for 10 min followed by various combinations of flourochrome-conjugated rat anti-mouse monoclonal antibodies ( BD-Pharmingen ) specific for CD45 ( 30-F11 ) , MHCII I-A/I-E ( M5/114 . 15 . 2 ) , CD11b ( M1/70 ) , CD11c ( HL3 ) , CD3 ( 17A2 ) , CD19 ( eBio1D3 ) , CD4 ( RM4-5 ) , CD8 ( 53–6 . 7 ) , CD44 ( IM7 ) , CD69 ( H1 . 2F3 ) , CD62L ( MEL-14 ) , B220 ( RA3-6B2 ) , podoplanin ( 8 . 1 . 1 ) , CD31 ( MEC13 . 3 ) for 35 min in FACS buffer ( 2% FCS in 5mM EDTA , 0 . 1% azide ) . For analysis FoxP3 ( FJK-16s ) and Ki67 ( SolA15 ) ( eBioscience ) FoxP3 and Ki67 staining set were used according to manufacturer’s instructions ( eBioscience ) . To detect cytokine production , cells were restimulated with PMA/Ionomycin ( Sigma ) and treated with Brefeldin A ( Sigma ) , prior to intracellular IFNγ ( XMG1 . 2 , BD Pharmingen ) and surface staining , as previously described [43] . Irrelevant isotype-matched antibodies were used to determine levels of non-specific binding . The sample were acquired on LSRII ( BD Bioscience ) and analysed by FlowJo version 10 ( Treestar ) , using the gating strategies shown in S4 Fig . For transfer of cells , superficial LNs from young adult ( <8 weeks ) mice were collected and total lymphocytes harvested by making single cell LN suspension as described above . To enable tracking of lymphocytes we either used congenic cell , GFP-expressing cell or cells labelled with CFSE as previously described [45] . FACS determined the composition of transferred cell . The number of cells and time of transfer are indicated in respective figures . For deworming , mice were treated with pyrantel pamoate per oral gavage 40mg/mice , 3 times with one-day interval . Efficacy of de-worming was checked by assessment of egg in faeces samples . De-worming was considered effective when no eggs were detected . Fingolimod treatment: 3 mg/kg FTY720 ( Sigma Aldrich ) were injected i . p . daily for total 28 days starting one day before infection with H . polygyrus . IL-7 treatment: Mice were injected i . p . with 4μg/mouse recombinant murine IL-7 ( R&D Systems , 407-ML/CF ) / every second day for 28 days , starting 4 hours after infection with H . polygyrus . Serum was separated from venous blood collected from mice with or without H . polygyrus infection was collected and stored at -80°C until assessment . Serum concentrations were determined using the IL-7 ELISA kit ( M7000 , R&D System ) following manufacturers’ instructions . Total RNA was extracted from inguinal LN using Trizol ( Thermo Scientific ) according to manufacturer’s instructions and the RNA concentration was determined by Nanodrop spectrometry ( NanoDrop 2000c , Thermo Scientific ) . For RNA sequencing the RNA quality was determined using a bio-analyzer ( CliperLabchip GX ) and three samples from each group with RIN values above 6 . 0 were selected for further processing . 2 μg of RNA was used to construct sequence library using Illumina TruSeq Stranded mRNA library Prep kit ( Illumina ) . Sequencing was then performed on a HiSeq 2500 ( Hiseq Control Software2 . 2 . 58/RTA 1 . 18 . 64 ) with a 2x 125 bp paired-end reads . Reads were mapped to the Mouse genome assembly , build NCBIM37 , with Tophat/2 . 0 . 4 , merged and duplicates were removed using picard-tools/1 . 29 in samtools . 5 . 3–8 . 3 million mappable reads were obtained from the samples . Gene constructs were generated using htseq/0 . 6 . 1 on bam files with duplicates included . Fragment Per Kilobase of transcript per Million mapped reads ( FPKM ) for genes and transcripts were generated using cufflinks/2 . 1 . 1 . Correlations within replicate groups were used to assess the background FPKM , which was set to 3 . A transcript was thereafter only considered if it passed the following criteria A ) evidence of the transcript ( FPKM > 3 ) in at least three of six animals , B ) not being of non-coding nature ( filtered using the available ncRNA information available in the ENSEMBL GRCm38 . p5 genome assembly ) . In the instance where a transcript displayed FPKM <3 in three animals or less , those values were adjusted to 3 . FPKM values were thereafter Log2 transformed and used in the make . heatmap1 function distributed in the R-project NeatMap package to generate non-clustered heatmaps with median row normalizations to visualize relative gene expression . Relative expression of selected genes is shown based on assumed association with inflammation , lymphocytes , cell migration , cell death/survival and extra cellular matrix , all deemed to be of possible relevance to outcome of skin BCG infection . Gene expression data is available at NCBI , SRA ( BioProject accession PRJNA433170 ) . For conventional RT-qPCR first strand cDNA generation was performed using Invitrogen Superscript . Real-time PCR was performed on CFX384 ( Bio-Rad ) using FAM-MGB labelled primer/probe for mRNA of interest ( all best coverage TaqMan gene expression assay , Applied Biosystems ) or SYBR green ( Sigma ) identification of double stranded DNA following amplification using cDNA specific primers ( S1 Table ) . Expression of HRPT was used as housekeeping ( hk ) reference and the relative expression of gene expression was calculated using the 2-ΔΔCt method . Popliteal lymph nodes were placed in OCT compound , frozen on dry ice , and stored in -20°C until sectioning . 8 μm thick sections were blocked for 1 hour in 2 . 5% normal goat serum ( Jackson ImmunoResearch ) followed by incubation with directly conjugated ( rat anti-CD3 AF647 clone 17A2 and rat anti-B220 AF488 clone RA3-6B2; Biolegend ) and unconjugated polyclonal rabbit anti-collagen IV ( Abcam ) antibodies in goat serum over night at 4°C . Sections were then incubated with secondary antibody ( chicken anti-rabbit AF594; Invitrogen ) for 1 hour in goat serum and mounted in mounting media with DAPI ( Vectachield , Vector Laboratories ) . Images were acquired with a confocal microscope ( Zeiss LMS 800 ) and analysed using Fiji ( ImageJ ) software . Statistical analyses were done using PRISM versions 6 and 7 ( GraphPad ) . Groups consisted of five or more mice at the start of an experiment . Samples/animals were excluded from the analysis if the worm infection was considered a failure ( i . e . worm burden in an untreated mouse at termination log10 less than normal , which typically equals≤10 worms per gut ) . In groups that received treatments , outliers were removed following Grubb’s test to define single outliers ( GraphPad online tools ) . All exclusion criteria were set before the start of the study . For comparison between two groups , we used student’s t-test , or non-parametric Mann-Whitney U test if samples diverged from normal distribution . Kruskall-Wallis test was used for multiple comparisons . Correlations were done using Pearson’s linear correlation test . Mean and SD are shown if not otherwise indicated . Statistically significant differences between groups are indicated as *p<0 . 05 , **p<0 . 01 , ***p<0 . 001 . Animal experiments were conducted in accordance to national regulations outlined in L150 ( Föreskrifter och allmänna råd om Försöksdjur , SJVFS 2012:26 ) . Animals were euthanized by cervical dislocation in accordance with approved ethical protocol . Isofluran anesthesia was used for footpad injections . Ethical approval was granted by the regional ethical board ( Stockholms djurförsöksetiska nämnd ) permit number N171/14 with amendment N131/16 and exemption from L150 Dnr 5 . 2 . 18-7344/14 approved by the Swedish Board of Agriculture .
|
Infections with intestinal nematodes may be one explanation to why BCG vaccination is less effective in areas of high worm burden . In support of this , we recently showed that chronic intestinal nematode infection resulted in reduced immune responses and higher mycobacterial burden at distal sites . How a gut-dwelling nematode modulate immune responses in skin-draining lymph nodes ( LN ) was not clear . We found a reduced expansion of LN draining the BCG injected footpad in worm-infected animals , but no evidence for a spread of regulatory cells or cytokines to the BCG-draining LN . Interestingly , we found that mice chronically infected with intestinal worms had significantly smaller skin-draining LN . We propose that the expansion of mesenteric lymph nodes ( mLN ) occur at the cost of other LN , leading to atrophy of skin-draining LN . Expansion of the lymphocyte pool by IL-7 , allowed worm-infected animals to maintain larger skin-draining LN while the mLN did not further expand . De-worming treatment of mice eventually restored the cellularity of skin-draining LN . This , however , took time indicating that effect of worms persisted long after the infection cleared . By de-worming and allowing time for the LN to recover , the cellular responses to BCG injection in the footpad were restored in the draining popliteal LN . Thus , paucity of lymphocytes at peripheral sites can explain impaired peripheral immune responses in worm-infected animals .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"blood",
"cells",
"t",
"helper",
"cells",
"dermatology",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immunology",
"parasitic",
"diseases",
"nematode",
"infections",
"skin",
"infections",
"cytotoxic",
"t",
"cells",
"digestive",
"system",
"infectious",
"diseases",
"white",
"blood",
"cells",
"animal",
"cells",
"t",
"cells",
"immune",
"response",
"lymphocytes",
"gastrointestinal",
"tract",
"cell",
"biology",
"anatomy",
"biology",
"and",
"life",
"sciences",
"cellular",
"types"
] |
2018
|
Atrophy of skin-draining lymph nodes predisposes for impaired immune responses to secondary infection in mice with chronic intestinal nematode infection
|
Genetic variation at immunoglobulin ( Ig ) gene variable regions in B-cells is created through a multi-step process involving deamination of cytosine bases by activation-induced cytidine deaminase ( AID ) and their subsequent mutagenic repair . To protect the genome from dangerous , potentially oncogenic effects of off-target mutations , both AID activity and mutagenic repair are targeted specifically to the Ig genes . However , the mechanisms of targeting are unknown and recent data have highlighted the role of regulating mutagenic repair to limit the accumulation of somatic mutations resulting from the more widely distributed AID-induced lesions to the Ig genes . Here we investigated the role of the DNA damage sensor poly- ( ADPribose ) -polymerase-1 ( PARP-1 ) in the repair of AID-induced DNA lesions . We show through sequencing of the diversifying Ig genes in PARP-1−/− DT40 B-cells that PARP-1 deficiency results in a marked reduction in gene conversion events and enhanced high-fidelity repair of AID-induced lesions at both Ig heavy and light chains . To further characterize the role of PARP-1 in the mutagenic repair of AID-induced lesions , we performed functional analyses comparing the role of engineered PARP-1 variants in high-fidelity repair of DNA damage induced by methyl methane sulfonate ( MMS ) and the mutagenic repair of lesions at the Ig genes induced by AID . This revealed a requirement for the previously uncharacterized BRCT domain of PARP-1 to reconstitute both gene conversion and a normal rate of somatic mutation at Ig genes , while being dispensable for the high-fidelity base excision repair . From these data we conclude that the BRCT domain of PARP-1 is required to initiate a significant proportion of the mutagenic repair specific to diversifying antibody genes . This role is distinct from the known roles of PARP-1 in high-fidelity DNA repair , suggesting that the PARP-1 BRCT domain has a specialized role in assembling mutagenic DNA repair complexes involved in antibody diversification .
The generation of high affinity antibodies through affinity maturation in B cells relies on the introduction of mutations into expressed immunoglobulin ( Ig ) gene alleles by somatic hypermutation ( SHM ) or gene conversion ( GCV ) . These closely related processes are mediated through introduction of a DNA lesion by activation-induced cytidine deaminase ( AID ) , followed by fixation of a mutation at or nearby the damage site via a mutagenic , rather than the usual conservative , DNA repair mechanism [1] , [2] . Mutations must be restricted to the Ig genes to protect the rest of the genome from accumulating potentially dangerous mutations , although this protection is far from perfect . Analysis of the mechanisms that direct mutagenesis to Ig loci has revealed the existence of multiple layers of regulation . One level of control is temporal regulation of expression of AID to activated B-cells in germinal centers , where cells with non-beneficial mutations can be quickly eliminated [3] . Another level of control is targeting of AID-mediated deamination to expressed Ig loci and , less frequently , a subset of other expressed genes through an as yet undefined transcription-dependent mechanism [4] , [5] . A third level of control is the Ig-specific targeting of mutagenic repair . While identical lesions at non-Ig loci are usually repaired by a high-fidelity mechanism , at Ig loci , a mutagenic repair pathway predominates , either through translesion synthesis by error-prone polymerases or GCV [6] . While mutagenesis is necessary for high affinity antibody production , mistargeting of either the AID-mediated deamination events or the mutagenic repair of incidental mutations has been linked to the generation of B-cell lymphomas and leukemias through the introduction of mutations into tumor suppressors and proto-oncogenes such as Bcl6 , Myc , RhoH , Pim1 , and Pax5 [7] , [8] , [9] . Recent data suggest that mistargeting of mutations occurs more frequently than previously thought , highlighting the importance of understanding how the processes that induce these mutations are targeted to specific genetic loci [6] , [10] , [11] . However , insights into the biochemistry through which either DNA lesions or mutagenic repair are targeted have been difficult to achieve , and so far have been limited to the definition of cis-acting DNA elements required for active mutagenesis at Ig loci [12] , [13] . The enzyme PARP-1acts as a gatekeeper of DNA repair . It is one of the first proteins to respond to DNA damage , where it binds and recruits the appropriate DNA repair enzymes . There is a slower , background level of repair in PARP-1 deficient cells , but DNA repair is severely impaired and these cells are rendered hypersensitive to DNA damaging agents such as methyl methane sulfonate ( MMS ) , N-Methyl-N′-Nitro-N-Nitrosoguanidine ( MNNG ) , and ionizing radiation [14] , [15] , [16] , [17] . In addition to a well-established role in base excision repair ( BER ) , there is evidence that suggests that PARP-1 may also play a role in repairing double strand breaks , although whether by homologous recombination ( HR ) , non-homologous end joining ( NHEJ ) , or micro-homology mediated end joining ( MMEJ ) is still the subject of lively debate [14] , [18] , [19] , [20] , [21] . A potential clue to the mechanisms of mutation targeting has been suggested by a recent report that the enzyme PARP-1 is constitutively bound to a DNA sequence within the Bcl-6 gene [22]—a locus which is frequently the subject of off-target mutations in B-cells [9] , [23] . This observation prompted us to evaluate mutation targeting to the Ig loci in a PARP-deficient variant of chicken DT40 B-cell line , in which Ig loci are constitutively mutated via GCV . Remarkably , we observe a nearly complete loss of GCV at Ig loci in PARP-deficient cells that is independent of the rate of AID-induced DNA lesioning . Functional analysis of PARP-1 variants demonstrated that PARP-1 is necessary for repair of AID-mediated DNA lesions , and that the capacity of PARP-1 to support GCV requires its BRCT domain which , to our knowledge , has no previously characterized function .
PARP-1 is thought to be one of the first proteins to respond to DNA strand breaks , where it binds and recruits the appropriate DNA repair enzymes [14] , [18] , [19] , [20] , [21] . Consistent with this model , PARP-1 deficient DT40 cells are rendered hypersensitive to DNA damaging agents such as MMS , MNNG , and ionizing radiation [14] , [15] , [16] , [17] . To further define the parameters of PARP-1's role in these repair processes , we evaluated the sensitivity of PARP-1−/− DT40 cells to MMS exposure , and assessed the capacity of human WT PARP-1 ( hPARP ) , a DNA binding mutant of hPARP-1 ( dZF2 ) , and two enzymatically inactive variants of hPARP-1 ( DBD-CAT and E988K ) to restore survival upon MMS challenge ( Figure 1A and B ) . As expected , hPARP expression fully restores MMS resistance to the PARP-1−/− cells , while the PARP-1 mutant lacking a DNA binding domain due to mutations in the cysteines critical for zinc finger folding is similar in phenotype to the knockout . In contrast , a catalytic inactive PARP-1 mutant that contains only the DNA binding domain and the catalytic domain minus the WGR portion exhibited poorer survival than the PARP-1−/− cells ( Figure 1B ) . As previously shown , this is likely due to the aggregation of inactive PARP-1 molecules at the site of DNA damage which could block access of DNA repair enzymes to the damaged site and/or deplete free PARP-1 and prevent binding and recruitment of repair enzymes to other sites of damage [28] . Surprisingly , we found that even after repeated attempts in parallel with successful transfection of other PARP-1 variants , we could not reconstitute PARP-1−/− cells with a full length , inactive variant of PARP-1 ( E988K ) , and thus we could not analyze the phenotype of PARP-1−/− cells reconstituted with that variant . As this variant had been successfully expressed in mouse embryonic fibroblasts previously [28] , we reasoned that a major difference between the DT40 cell context and other cell lines is the presence of constitutive AID expression and diversification of the Ig loci in DT40s . We therefore attempted expression of PARP-1 ( E988K ) in both wild type and AID-deficient DT40 cells . Remarkably , we were unable to grow out any resistant clones in wild type DT40 cells , but clones stably expressing E988K readily grew from the parallel transfection of AID−/− DT40s , with the transfection yielding dozens of transformants of which 15 were subjected to further analysis to verify expression ( Figure 1C shows a representative clone ) . When we subsequently attempted to reconstitute E988K , AID−/− DT40 clones with AID in the 4/TO vector ( Invitrogen ) with zeocin selection , of the 4 clones which eventually grew in the selective media , all had downregulated expression of E988K PARP-1 to below our limits of detection ( as detected by Western blot , unpublished data ) . This dramatic selection against clones that coexpress AID and E988K suggested that PARP-1 plays an important role in the repair of lesions induced by AID . The sensitivity of DT40s expressing AID to expression of PARP-1 E988K suggested that PARP-1 has a requisite role in repair of AID-induced lesions , either at mutating Ig loci or off-target lesions genome-wide . To assess the function of PARP-1 at mutating Ig loci , we sequenced the variable regions of the Ig light and heavy chains in PARP-1−/− and WT cells . While mutations still detectably accumulated within Ig genes , although at a dramatically reduced rate , the PARP-1−/− cells had essentially no discernible GCV events ( p< . 0001 at IgL and p = . 0436 at IgH , Figure 2A and B ) . To confirm that the observed defect in GCV was due to the actions of PARP-1 , we examined a cell line reconstituted with human PARP-1 ( hPARP ) ( Figure 2C ) and found that hPARP restored the GCV frequencies to wild type levels or above at both IgL ( p = . 0001 ) and IgH ( p = . 0108 ) ( Figure 2A and B ) , suggesting that PARP-1 not only is required to repair AID-mediated deaminations but also influences the outcome of the resulting repair . In this analysis , all mutations that matched chicken Ig pseudogene sequences published in the NCBI public database were counted as GCV events . This includes a subset of “ambiguous” mutations , which match the pseudogenes but occur as a single nucleotide change such that we cannot exclude the possibility that it arose as a point mutation . While the work of Saberi et al . shows that these mutations are generally true GCV events [29] , to ensure that the process of classifying GCV events is not affecting our conclusions , we have also analyzed the data with the “ambiguous” mutations excluded from the analysis or scored as point mutations ( Figure S2 ) . In these additional analyses , the difference in GCV frequencies between WT and PARP−/− cells at the IgL locus remains highly significant ( p< . 0001 and p = . 0006 , respectively ) , although the number of mutations scored at the less well defined IgH locus was not sufficient to reveal a significant difference in these more conservative analyses ( p = . 1023 and p = . 1280 , respectively ) . Subsequent restoration PARP-1 expression restores GCV at both IgL ( p = . 0002 and p = . 0029 ) and IgH ( p = . 0108 and p = . 0108 ) . The individual mutations observed at IgL can be found in Figure S1 and a schematic of mutations observed at IgH is shown in Figure S2E . In the course of characterizing the GCV of PARP-1−/− cells , we observed that the PARP-1−/− cell line and its hPARP derivative express less AID than the WT DT40 cells , and that the PARP-1−/− and hPARP lines had a lower overall mutation rate . As low AID expression seemed likely to be the inadvertent result of selection during derivation of the parental PARP-1−/− clone that was subsequently carried over to the reconstituted hPARP cell line , we investigated whether the reduction in AID expression and corresponding overall mutation rate could be influencing the proportion of GCV events . For this purpose , we used a retroviral vector to overexpress gallus gallus AID in each cell line , and matched AID expression as well as IgL transcript levels in selected clones ( Figure 3B and C ) , as the rate of target gene transcription has also been shown to affect mutation rate [30] , [31] , [32] . Consistent with previous reports , overexpressing AID does not significantly increase the proportion of mutations which are GCV events [29] and does not restore GCV to the PARP-1−/− cells , in spite of increasing the AID expression to well above the original WT levels ( Figure 3A and D ) . Interestingly , overexpression of AID does not significantly increase the mutation rate in the PARP-1−/− cells as it does in hPARP , and may be revealing a dose-dependent effect of PARP-1's ability to mediate GCV repair of deamination events as increasing AID expression results in a WT GCV phenotype intermediate to PARP-1−/− and hPARP ( Figure 3A ) . Furthermore , our ability to generate high stable expression of AID in PARP-1−/− cells suggests that the reduced mutation rate in PARP-1−/− cells relative to hPARP reconstituted cells cannot be explained by loss of cells that sustain AID-mediated lesions , but rather may reflect a reduced rate of mutagenic repair of AID-mediated lesions . As an initial step towards defining the role of PARP-1 in repair of AID-induced lesions , we investigated the influence of chromatin accessibility on mutagenic repair at Ig loci . It has been proposed that PARP-1 modifies histones at the site of DNA damage to open chromatin and increase the accessibility of a damaged site to repair enzymes [33] . To determine whether altered chromatin accessibility could account for deficient GCV in PARP-1−/− cells , we cultured PARP-1−/− cells in trichostatin A , a histone deacetylase ( HDAC ) inhibitor , as this has previously been shown to increase GCV rates , presumably via increasing accessibility of the pseudogene repair templates to repair machinery [34] . While the percent and length of GCV events in WT cells went up dramatically , indicating that the treatment was effective ( Figure 4A ) , the PARP-1−/− cells are still unable to carry out GCV ( Figure 4B ) . Thus , the mechanism of PARP-1's role in promoting GCV at Ig loci is not through increasing chromatin accessibility . Ig GCV in DT40s proceeds through a pathway involving HR . Since PARP-1 has previously been implicated in HR repair , we assessed whether PARP-1's capacity to promote Ig GCV was a part of a general role for PARP-1 in promoting HR by evaluating the capacity of PARP-1−/− cells to mediate HR in response to a single DNA double strand break generated by the homing endonuclease I-SceI ( Figure 5A ) . In a single-copy , integrated assay of HR , we find that the expression of PARP-1 does not promote HR . Rather , in agreement with studies by Wang et al . , PARP-1 may suppress HR and promote alternate DNA repair pathways ( Figure 5B ) [19] . This result indicates that the activity of PARP-1 at Ig loci is not part of a global role promoting HR , and further supports the hypothesis that PARP-1 has a specific role within GCV as a mutagenic repair pathway operating at Ig loci . It also raises the question of whether it is best to consider GCV at Ig loci as an unusual , mutagenic instance of otherwise high-fidelity HR or if a better model would be to consider GCV one pathway for mutagenic repair at Ig loci that uses much , but not all , of the same repair machinery as HR . To further investigate the mechanism by which PARP-1 mediates GCV , we evaluated the role of specific PARP-1 subdomains in promoting mutagenic repair at Ig loci . Hypothesizing that the role of PARP-1 in mutagenic repair at Ig loci may be distinct from its established role in high-fidelity BER , we compared DNA BER and Ig GCV among PARP-1−/− clones reconstituted with PARP-1 variants containing deletions or inactivating mutations in each domain . Consistent with their lack of ability to reconstitute any detectable DNA BER ( see Figure 1 ) , sequence analysis of Ig loci in PARP-1−/− cells reconstituted with the dZF2 or DBD-CAT variants indicated that expression of either of these PARP-1 variants was unable to reconstitute Ig GCV ( Figure S3 ) . We then analyzed PARP-1 variants that retained significant capacity to reconstitute BER in the PARP-1−/− cells . Analysis of PARP-1−/− cells reconstituted with variants of PARP-1 containing either a deletion of the BRCT domain ( dBRCT lacks aa384–479 ) or the automodification domain sparing the BRCT domain ( dAMD lacks aa372–383 and 480–524 ) ( Figure 6A ) gave surprising results . In an MMS survival assay , the dBRCT cells survived as well as WT cells ( Figure 6C ) , consistent with previous data from our lab and others indicating that the BRCT portion of the automodification domain is not required for PARP activation via base-damaging agents or PARP-dependent base excision repair [16] , [35] . In the same assay , dAMD-expressing cells exhibited an intermediate survival phenotype when treated with MMS ( Figure 6C ) , consistent with the dAMD mutant exhibiting a significant but delayed capacity to catalyze NAD degradation and ADPR production in response to base damaging agents [16] . We then assayed the ability of these cell lines to undergo GCV at IgL and found that the capacity to mediate GCV does not correspond with the ability of the cells to carry out global high-fidelity DNA repair . In contrast , sequence analysis of Ig loci revealed that the dBRCT cells , like PARP-1−/− cells , were essentially unable to mediate GCV and had a correspondingly low overall mutation rate , whereas the dAMD cells were able to gene convert at IgL at or near WT levels ( Figure 6D and Figure S3 ) . This observation demonstrates that the role of PARP-1 in mutagenic GCV at Ig loci is distinct from its role in high-fidelity DNA BER . Furthermore , it implies a novel role for the BRCT domain of PARP-1 in mediating mutagenic DNA repair at Ig loci . When AID is overexpressed in cells overexpressing PARP-1 ( hPARP ) , we observe that the mutation rate increases accordingly . However , when AID is overexpressed in PARP-1−/− cells , we do not see a significant change in mutation rate . This suggests that while AID deamination is the initial rate-limiting step for SHM and GCV , expression of PARP-1 may further limit mutation rate during repair of these lesions ( Figure 7A and B ) . One potential explanation for these observations is that lack of PARP-1 selects against PARP-1−/− cells with high AID expression that would have accumulated mutations . To test this possibility , we evaluated the stability of AID overexpression over the 3 wk culture period used for accumulation of mutations , based on the rationale that selection pressure against mutations would lead to a decreased expression of AID . In this experiment , comparison of AID expression before and after 3 wk of culture revealed no detectable difference in AID expression , indicating that selection against the PARP-1−/− cells that would have accumulated mutations cannot account for the observed decrease in mutation rate in the PARP-1−/− cells ( Figure 7C ) . Thus , the observed lower mutation rate reflects either a decreased rate of DNA lesioning or an increased rate of faithful repair of AID-mediated DNA lesions . While the combined evidence of an established role for PARP-1 in DNA repair and the baseline level of mutations which continue to accumulate in the PARP-1−/− cells lead us to prefer the latter hypothesis , we sought to resolve this issue by expressing an UNG inhibitor UGI in the PARP-1−/− DT40 cells expressing endogenous levels of AID . UGI has been shown to reveal an “AID footprint” of activity by blocking BER of deaminated cytosines and increasing the relative frequency of mismatch repair mutations at A/T and “replication over” events , fixing C to T and G to A mutations at the site of deamination [36] . Correspondingly , we found that UGI expression results in decreased transversion mutations at G/C base pairs , and a relative increase in replication over errors at G/C and mismatch repair mediated mutations at A/T base pairs ( Figure 7E ) . Analyzing the rate of mutation by sequence analysis , we observed that UGI expressing PARP-1−/− cells exhibited an increased rate of mutation relative to their parent cell line , and now matched the mutation rate observed in the hPARP cells ( Figure 7D ) , chosen as a control because they express a similar level of AID ( see Figure 2 ) . These findings support our hypothesis that the low mutation rate of the parental cell line was not the result of decreased deamination events , but rather reflected an increased proportion of high-fidelity repair . This collection of evidence leads us to conclude that PARP-1 is promoting mutagenic repair at Ig loci and that AID-induced lesions are more likely to be repaired faithfully in the PARP-1−/− cells .
In this paper , we studied the role of PARP-1 in mediating repair of AID-induced lesions at diversifying Ig loci . By sequencing Ig light and heavy chain genes in PARP-1−/− DT40 B-lymphocytes , we found that the overall mutation rate is reduced and GCV is essentially eliminated in the absence of PARP-1 , and these defects can be fully reconstituted by expression of the hPARP-1 gene . Dissection of the biochemical mechanisms underlying PARP-1's involvement in Ig diversification demonstrated that PARP-1 DNA binding and BRCT protein-protein interaction domain are required , while the major site of automodification is not . Furthermore , while the overall mutation rate in PARP-deficient cells could be slightly increased by AID expression , restoration to WT levels required concomitant inhibition of uracil-DNA-glycosylase , suggesting that PARP-deficiency leads to an increased rate of high-fidelity repair at Ig loci through the UNG-dependent BER pathway . Taken together , our data suggest that PARP-1 is an important part of the biochemical processes that promote mutagenic repair over faithful repair at Ig loci , through a mechanism that requires an intact BRCT domain . The requirement for the BRCT domain of PARP-1 in promoting mutagenic repair at Ig loci revealed in our studies defines a novel role for the BRCT domain of PARP-1 . In spite of the extensive work which has lead to our current understanding of the subdomains of PARP-1 , the function of the BRCT portion of the automodification domain has remained a mystery , as it is not required for PARP-1 mediated repair of the other types of DNA damage which have been studied [16] , [35] . As BRCT domains are thought to function as protein-protein interaction domains , an intriguing possibility arising from our observations is that there may be a protein interaction partner of the PARP-1 BRCT domain which is involved in the mutagenic repair of AID-induced lesions but not involved in high-fidelity repair of other types of DNA damage . Identification of the hypothetical interaction partner ( s ) for the PARP-1 BRCT domain should further illuminate the mechanisms involved in PARP-1-dependent targeting and regulation of mutagenic DNA repair . PARP-1-dependent targeting of mutagenic repair may also account for the frequent mistargeting of mutations to the Bcl-6 gene in B-cells , as PARP is constitutively targeted to the Bcl-6 locus via sequence specific binding of its zinc fingers [37] , and it would be interesting to see if PARP-1 also binds specifically to other common sites of mistargeted mutations involved in malignant transformation such as Bcl-2 , Pim1 , Pax5 , Myc , or RhoH . Diverse lines of evidence have recently developed to support the concept that AID-mediated DNA lesions are not uniquely targeted to Ig loci [10] , [11] but rather that off-target deamination events in germinal center B-cells went undetected because those that occur outside the Ig loci primarily undergo high-fidelity repair [6] . This new information emphasizes the importance of understanding not just how AID is targeted to Ig loci but , equally as important , how mutagenic repair of deaminations is targeted to the Ig loci to protect B cells from the dangers associated with antibody diversification such as oncogenesis . Our data demonstrating a role for PARP-1 in both high-fidelity BER genomewide and mutagenic repair of deaminations at Ig loci present a mechanism for the targeting of mutagenic repair . PARP-1 , known to play a key role in directing repair of alkylated DNA bases towards BER through interactions with the PARylated AMD , also directs repair down a mutagenic pathway at Ig loci through interactions with the BRCT domain . Our data suggest that in the WT system , PARP may be acting at the site of a single strand break generated by AP lyase to either ( 1 ) directly promote GCV repair of breaks that would otherwise undergo high-fidelity repair or to ( 2 ) prevent high-fidelity repair at Ig loci , resulting in diversion to the less efficient error-prone repair pathways , including GCV and SHM ( Figure 8 ) . When normal GCV pathways are blocked , such as in XRCC2/3 knockouts , the lesions are diverted to other mutagenic repair pathways , such that the total mutation rate is unchanged [38] . In PARP-1−/− cells , the lesions enter a high-fidelity repair pathway rather than undergoing GCV , which is more consistent with the latter hypothesis . On the other hand , an early report from PARP-1−/− mice would favor the former hypothesis , as the authors failed to find a hypermutation defect in B cells from those mice [27] . However , in those experiments Jacobs et al . examined ex vivo germinal center B cells which had undergone intense selective pressure for the accumulation of mutations . As we have shown that SHM is not eliminated but reduced in the absence of PARP , and the reported number of mutations in the PARP-1−/− mice was reduced by more than one third ( although they did not report how many sequences were analyzed from each mouse ) , our results are not in contradiction to this early report , and PARP-1 could still play a role in inhibiting high-fidelity repair during SHM and GCV . Additionally , a subsequent report identified reduced T cell-dependent responses and reduced AID expression in PARP-1−/− mice [26] , which could be explained by increased cell death among hypermutating B cells consistent with the theory that PARP-1−/− B cells accumulate fewer mutations than WT B cells , and that the cells which fail to increase affinity by hypermutation are selected against in germinal centers . Our results establish a role for PARP-1 in the repair of the AID-induced lesions required for SHM and GCV , and further show that this role is mediated in part by the BRCT domain of PARP-1 , a domain with , to our knowledge , no previously known function . Without the BRCT domain of PARP-1 , the mutation rate is lower and GCV essentially absent in DT40 B cells . The requirement for the BRCT domain of PARP-1 may suggest that an interaction partner for PARP-1 is important for mediating this role , likely by inhibiting high-fidelity repair in the UNG-dependent BER pathway and allowing alternative , mutagenic repair pathways to predominate or , alternatively , by directly promoting GCV , thus allowing fewer lesions to enter a high-fidelity repair pathway .
The variable regions of IgL and IgH were amplified with Accuprime Pfx and blunt end Topo cloned . Single colonies were picked and sequenced using the M13 reverse primer . The primers used for PCR were IgL-F: CAGGAGCTCGCGGGGCCGTCACTGATTGCCG , IgL-R: GCGCAAGCTTCCCCAGCCTGCCGCCAAGTCCAAG , IgH-F: CGGGAGCTCCGTCAGCGCTCTCTGTCC , IgH-R: GGGGTACCCGGAGGAGACGATGACTTCGG . A baseline rate of mutation was determined by sequencing an irrelevant gene , the constant region of IgL , or the variable region of IgL in AID−/− cells . The polymerases Pfx ( Invitrogen ) and Pfu Ultra ( Stratagene ) were also compared ( Figure S4 ) . The baseline mutation rates were well below the observed mutation rates in this study , and we decide to proceed with our analyses using Pfx Accuprime polymerase . Sequences were aligned using Phred and Phrap and viewed in Consed . High quality base discrepancies were noted and subjected to further analysis . As the total mutation rate was much lower than 1 mutation/read , tracks of multiple mutations in a read were scored as GCV events . Single mutations for which no donor template could be identified were scored as point mutations . To categorize ambiguous mutations ( which match the pseudogene templates but occur in isolation ) , results were compared when these mutations were ( 1 ) excluded from the analysis , ( 2 ) always considered point mutations , and ( 3 ) always considered GCV events . These changes made little difference to the final analysis as the mutations in the PARP-1−/− and dBRCT cell lines very rarely matched the pseudogene sequences through either blast searches or direct comparison to a database of collected pseudogene sequences and so were able to be clearly scored as point mutations . While the ability of WT DT40s to undergo GCV is not in question , in order to avoid missing any GCV events which may occur in the PARP−/− and dBRCT cell lines , it was decided to use the most inclusive definition of a GCV event , which is every mutation that matches the pseudogene sequences ( annotated as “similar to immunoglobulin lambda chain” within gi 118098819 ) by blastn . At IgH , where the pseudogenes are not well characterized , a match resulting from a whole genome blast which was located adjacent to IgH was considered a suitable donor sequence for a GCV event and again , and all mutations with a donor template match were scored as GCV events . In spite of the poor assembly at IgH , the availability of donor templates as assessed is equivalent for all the cell lines used , so they may be compared . p values were generated using Fisher's exact test . Q-PCR was performed on a BioRad icycler using the BioRad SYBR green master mix . The annealing temperature was 58°C . The following primers were used: IgL-F: caggagctcgcggggccgtcactgattgccg , IgL-R: gcgcaagcttccccagcctgccgccaagtccaag , Beta Actin F: tgagagggaaatcgtgcgtgacatc , Beta Actin R: caggaaagagggttggaacagagcc . IgL transcript level was normalized to β-actin and hPARP expression using the ΔΔCt method . Data analysis was performed using Microsoft Excel and Graphpad Prism . Whole cell lysates were separated by SDS-PAGE , transferred to Millipore immobilon membrane , and probed with the following antibodies: PARP-1: ALX-210-302 ( Alexis Biochemicals ) , βActin: A2228 ( Sigma-Aldrich ) , and AID: LS-C34861 ( Lifespan Biosciences ) . Secondary antibodies were labeled with IRdye-700CW or IRdye-800CW and analysis and quantification was done on the LICOR Odyssey Infrared Imager . DT40 cells were cultured in RPMI with 10% FBS , 5% CS , Pen/Strep , and b-ME at 41° . PARP-1−/− cells and the WT parent cell line were a generous gift from S . Takeda . Cell lines reconstituted with hPARP-1 and the variant hPARP constructs were generated by electroporation of PARP-1 cDNA in the 5/TO vector ( Invitrogen ) at 550V , 25 uF in 4 mm cuvettes using the GenePulser from Bio-Rad . Colonies which grew in hygromycin were matched for PARP-1 expression by Western blot . AID−/− cells were a gift from JM Buerstedde . AID−/− cells transfected with E988K were then transfected with human AID cDNA in the 4/TO vector ( Invitrogen ) as above , with selection in zeocin . Cells treated with TSA were incubated in 2 ng/mL TSA ( Sigma T8552 ) , refreshed daily , for the entire culture period . Cell lines overexpressing AID used for mutation analysis were generated by retroviral transduction of a plasmid encoding chicken AID IRES GFP , another gift from S . Takeda . GFP high cells were subcloned and AID expression was measured by Western blot . All cells used for mutation analysis were subcloned by limiting dilution immediately before the culture period to ensure a homogenous starting population . Cells were then allowed to accumulate mutations for a period of 6 wk in initial experiments and 4 wk for experiments in which all the cell lines overexpress AID . There were no notable differences in generation time and all cells were split 1∶16 every other day . For calculations which include generation time , 12 h was used . A recombination substrate encoding BFP containing an intron with the recognition sequence for I-SceI was transduced into DT40s using a limiting titer of lentivirus to bias toward single integration events ( cultures with less than 5% transduction efficiencies were used ) . Transient transfection of a GFP repair template plus I-SceI expression plasmid was performed by electroporation at 250 V , 950 uF in 4 mm cuvettes . Parallel transfection of a GFP control plasmid was used to estimate transfection efficiency and frequency of HR was calculated as the percent of mCherry positive cells that were also GFP positive , divided by the percent that were positive for the GFP control . Cells were exposed to MMS at the indicated concentrations for 1 h at 37° . They were then washed 2× in fresh media and resuspended in 3 mL media . 450 ul of 3% agar was added and 1 mL was plated in triplicate . Plates were grown for 3–4 d at 41° before colonies were counted .
|
To produce a limitless diversity of antibodies within the constraints of a finite genome , activated B cells introduce random mutations into antibody genes through a process of targeted DNA damage and subsequent mutagenic repair . At the same time , the rest of the genome must be protected from mutagenesis to prevent off-target mutations which can lead to the development of lymphoma or leukemia . How antibody genes are specifically targeted is still largely unknown . A potential player in this process is the DNA-damage-sensing enzyme PARP-1 , which recruits DNA repair enzymes to sites of damage . Using a chicken B cell lymphoma cell line because it has only a single PARP isoform and constitutively mutates its antibody genes , we compared the types of mutations accumulated in PARP-1−/− cells to wild type . We found that in cells lacking PARP-1 , the major pathway of mutagenic repair was disrupted and fewer mutations than normal were introduced into their antibody genes . To identify what might be important for mutagenesis , we tested different factors for their ability to rescue this mutagenic deficiency and found a role for the BRCT ( BRCA1 C-terminal ) domain of PARP-1 , a consensus protein domain known to be involved in directing protein-protein interactions . Our evidence suggests that PARP-1 may be interacting with another hypothetical protein via its BRCT domain that is required for the mutagenic rather than faithful repair of DNA lesions in the antibody genes .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"immunology/genetics",
"of",
"the",
"immune",
"system",
"biochemistry/replication",
"and",
"repair",
"biochemistry"
] |
2010
|
The BRCT Domain of PARP-1 Is Required for Immunoglobulin Gene Conversion
|
The ability to predict the response of a cancer patient to a therapeutic agent is a major goal in modern oncology that should ultimately lead to personalized treatment . Existing approaches to predicting drug sensitivity rely primarily on profiling of cancer cell line panels that have been treated with different drugs and selecting genomic or functional genomic features to regress or classify the drug response . Here , we propose a dual-layer integrated cell line-drug network model , which uses both cell line similarity network ( CSN ) data and drug similarity network ( DSN ) data to predict the drug response of a given cell line using a weighted model . Using the Cancer Cell Line Encyclopedia ( CCLE ) and Cancer Genome Project ( CGP ) studies as benchmark datasets , our single-layer model with CSN or DSN and only a single parameter achieved a prediction performance comparable to the previously generated elastic net model . When using the dual-layer model integrating both CSN and DSN , our predicted response reached a 0 . 6 Pearson correlation coefficient with observed responses for most drugs , which is significantly better than the previous results using the elastic net model . We have also applied the dual-layer cell line-drug integrated network model to fill in the missing drug response values in the CGP dataset . Even though the dual-layer integrated cell line-drug network model does not specifically model mutation information , it correctly predicted that BRAF mutant cell lines would be more sensitive than BRAF wild-type cell lines to three MEK1/2 inhibitors tested .
Over the past two decades , substantial improvements in high-throughput profiling technologies and systems approaches have increased expectations that personalized or precision medicine will become the paradigm of future medical science [1–3] . In contrast to the one-size-fits-all approach that has dominated cytotoxic chemotherapy , personalized medicine exploits tumor response and vulnerability based on identified molecular traits to overcome some of the limitations associated with conventional symptoms-oriented disease diagnoses and therapies . The most important step in implementing personalized medicine will be the identification of biomarkers useful for predicting the drug response of a given patient [4–6] . However , the development of predictive biomarkers would require substantial efforts and is often prohibitively expensive in human or animal models . Therefore , many studies conduct large-scale drug screenings on cultured human cell line panels to identify predictive biomarkers [7] . One of the earliest such attempts is the NCI-60 study [8 , 9] , which included a set of 60 human cell lines and their responses to more than 100 , 000 chemical compounds . Drug response results for the NCI-60 dataset [10 , 11] revealed that different types of cancers have different drug response signatures , and that different tumors derived from the same type of cancer may have distinct molecular patterns [12] . Two recent consortiums , the Cancer Cell Line Encyclopedia ( CCLE ) [13] and Cancer Genome Project ( CGP ) [14] , systematically addressed the issue of predictive biomarker identification by collectively analyzing around 1 , 000 clinically-relevant human cell lines and their pharmacological profiles for 149 cancer drugs . These two studies also included the gene expression profiles and mutation status for each cell line , and applied the elastic net model to select expression and mutation signatures that are predictive of drug responses . Based on the same dataset , Geeleher et al . applied another sparse regression model , Ridge , to predict drug response for breast cancer cell lines using baseline gene expression data [15] . Brubaker et al . used a probabilistic graphical model named PARADIGM to infer patient-specific genetic activity by integrating copy number and gene expression data into a factor graph model [16] . Menden et al . integrated genomic features of cell lines ( mutation , copy number and microsatellite instability ) with chemical properties of drugs to represent each cell line-drug pair , and used neural network to predict drug response in CGP dataset [17] . Ammad-ud-din et al . proposed a kernelized Bayesian matrix factorization model to integrated drug property matrix and cell line genomic properties matrix [18] . This kind of approach could capture the nonlinear relationships between drug response and chemical descriptors ( or cell line genomic features ) by a kernel strategy and thus was adopted in many other areas including drug-target interaction prediction [19] . Despite achieving promising results for certain drugs , these approaches do not take into consideration two important characteristics of a cancer cell drug response screens , including: 1 ) genetically similar cell lines or samples may also respond very similarly to a given drug; and 2 ) structurally related drugs may have similar therapeutic effects due to their shared molecular structure or targeting patterns . Incorporating similarities between cell lines and drugs could potentially improve the drug response prediction . In this study , motivated by the integrated model in disease genes prioritization [20 , 21] , we constructed a dual-layer integrated cell line-drug network , and modeled the integrated similarity between cell lines based on their gene expression profiles , and between drugs based on their 1-D and 2-D chemical structures . We predicted the response of a given cell line to a drug based on a weighted model using either one or both layers of the cell line similarity network ( CSN ) and the drug similarity network ( DSN ) . Our proposed dual-layer integrated cell line-drug network model combines the predictions from the individual CSN and DSN layers , and predicts a response of a cell line to a drug based on how similar cell lines ( CSN ) respond to similar drugs ( DSN ) . Instead of selecting a large number of genomic features from previous studies to predict tumor drug responses , our model uses only three parameters to build a prediction model to decrease the risk of overfitting . Using CCLE and CGP studies as benchmark datasets , we evaluated the predictive power of our model and found that our dual-layer integrated cell line-drug network model is significantly better than model that use either the CSN or the DSN layer alone , as well as the elastic net model . We also applied the dual-layer network model to fill in all of the missing drug response values ( activity area and IC50 ) in the CGP dataset , and found that our predicted responses to three MEK1/2 inhibitors in the CGP study show a distribution very similar to other available drug response values .
We built the dual-layer integrated cell line-drug network model using large pharmacogenomics datasets from the CCLE [13] and CGP [14] studies . Experimentally determined drug responses , also referred to as drug sensitivities , which were measured as activity area and IC50 in both studies . Notably , a higher value of activity area or lower value of IC50 indicates a better sensitivity of a cell line to a given drug . We first took activity area as drug response measurement . The CCLE study , for example , contains expression profiles of 491 cancer cell lines , as well as their response to 24 drugs . Since different drugs have different baseline values and ranges , we normalized the drug response data so that different drugs have the same baseline value and range across all cell lines . We calculated the Pearson correlation of gene expression profiles and the Pearson correlation of drug responses by activity area for each cell line pair . Drug sensitivity correlations were significantly higher for cell lines with more similar gene expression profiles ( Fig 1A ) . The CGP dataset contains response data and expression profiles for 653 cell lines treated with 139 drugs . In agreement with the CCLE observations , CGP cell lines with higher gene expression similarity show higher drug response correlations for all the drugs tested ( Fig 1B ) . These results suggest that cell lines with similar gene expression profiles exhibit similar drug responses . Seeing that similar cells exhibit similar drug responses , we next examined whether similar drugs have similar effects on cells . We downloaded the chemical structure files from PubChem [22] for the drugs used in the CCLE and CGP datasets , extracted the 1-D and 2-D structural features of each drug using PaDEL [23] , and calculated the Pearson correlation between two drugs using these structural features . Here , 1D descriptors refer to compositional or constitutional molecular properties , and 2D descriptors include different quantitative properties of the molecular topology ( see Materials and Methods for details ) . We divided all drug pairs into three groups according to their pairwise chemical structural similarities: low , intermediate and high . Drug pairs with more similar structures have significantly higher drug sensitivity correlations by activity area in both the CCLE ( Fig 1C ) and CGP ( Fig 1D ) datasets . This finding suggests that drugs with similar chemical structures show similar inhibitory effects across the cell lines tested . When using IC50 as drug response measurement , we observed a quite similar phenomenon in both datasets ( S1 Fig ) . A recent report showed inconsistent results between the CCLE and CGP datasets [24] . Our analysis above found the CSN and DSN hypotheses to be generally valid for both datasets , but the trend is a little stronger by using activity area than IC50 . This might be because the activity area is measured over the whole dose-response curve , which better captures the drug effects and cell responses . In contrast , the IC50 measurement , i . e . , the concentration at which the drug response reaches an absolute inhibition of 50% . So IC50 measurements only consider a single point on the dose-response curve to determine drug sensitivity for the cell lines , which might be noisier . Based on the above results , we developed a dual-layer integrated cell line-drug network model to predict anticancer drug sensitivity using existing cancer cell line expression profiles and drug response data ( Fig 2 ) . The model integrated three types of data: 1 ) gene expression profiles for each cell line; 2 ) 1-D and 2-D chemical structural properties of each drug; and 3 ) the drug response for each cell line . The top layer of the network , termed cell line similarity network ( CSN ) , predicts the response of cell line C to a given drug D using a linear model weighting drug response from cell lines with similar gene expression profiles to C . We calculated the gene expression correlations between C with all other cell lines ( Fig 2A and 2B ) , and gave higher weights to more similar ones . The bottom layer , termed drug similarity network ( DSN ) , predicts the response of cell line C to drug D weighting the response data on drugs similar to D in their chemical structures . We calculated the correlation of every drug pair based on their 1-D and 2-D chemical structure features in PubChem . The two layers , CSN and DSN , were connected using drug response data for the cell lines , which were represented as activity areas in CCLE [13] and CGP studies [14] . Notably , the network is not a complete bipartite graph , as some drug response data for some cell lines are missing in these studies , especially for the CGP dataset . Our model contains only three parameters , which determine how to weigh the different cell lines wC , how to weigh the different drugs wD , and how to combine them λ , respectively . wC can in turn to be written as wC ( C , Ci ) =e−[1−ρC ( C , Ci ) ]22σ2 , where σ determines the decay rate when cell expression correlation decrease . Within the σ range of [0 , 1] at 0 . 001 increments , we examined the top layer CSN and identified the optimal σ = 0 . 030 that minimizes sum of squares in prediction errors of all drugs in the CCLE dataset ( Fig 3A ) . wD can also in turn be written as wD ( D , Dj ) =e−[1−ρD ( D , Dj ) ]22τ2 , where τ determines the decay rate when drug similarity correlations decrease . Within the τ range of [0 , 1] at 0 . 01 increments , we examined the bottom layer DSN and identified the optimal τ = 0 . 40 that minimizes sum of squares in prediction errors among all cell lines in the CCLE dataset ( Fig 3B ) . Once σ ( hence wc ) and τ ( hence wD ) were determined , we calculated λ for each drug ( Fig 3C ) , the parameter weighting the relative contributions of CSN and DSN , to minimize the sum of the squared errors . Most of the drugs have λ larger than 0 . 5 ( Fig 3C ) , suggesting that DSN is more informative for predicting the drug response than CSN . The final prediction model for each drug contained the three optimized parameters . In order to test the robustness of the estimated parameters in our dual-layer integrated cell line-drug network model , we conducted leave-one-out cross-validation by singling out each drug-cell line pair as the test dataset . Performance was measured by the Pearson correlation between predicted and observed drug responses . We first applied our method to the CCLE dataset , and compared the results to those from elastic net regression as reported in the original study ( Fig 4A ) . Consistent with the above finding that λ is greater than 0 . 5 for most drugs , predictions based on DSN alone were better than those based on CSN alone for most drugs . In addition , the dual-layer model integrating DSN and CSN gave superior performance to either CSN alone or DSN alone , indicating that more information ( genomic features of cell lines and chemical features of drugs ) are helpful in drug response prediction . Three drugs , PD-0332991 , Irinotecan , and Nutlin-3 , showed similar performance between the dual-layer integrated cell line-drug network model and DSN alone , since their λ parameters are close to 1 . More importantly , our predictions using the dual-layer network model were significantly better than the previously published predictions using the elastic net model for most drugs , except Nilotinib . Notably , the drugs AZD530 and Nutlin-3 had poor prediction correlations using the elastic net model ( < 0 . 2 ) , although both were well predicted by our dual-layer integrated cell line-drug network model ( > 0 . 6 ) . The overall prediction performance of the dual-layer integrated cell line-drug network model across all the drugs was significantly higher than that of the elastic net model for the CCLE dataset ( Fig 4B ) . We believe that the improved performance of our model is partially due to the incorporation of drug chemical features , which are not used in the previous elastic net model . The scatter plots of observed versus predicted responses for 4 example drugs ( Fig 4C ) indicate that the good correlations did not arise from a small number of outliers ( the remaining 19 drugs are listed in S2 Fig ) . We also conducted leave-one-out cross-validation to evaluate the performance of the dual-layer integrated cell line-drug network model in the CGP study with activity area as drug response measurement . For drugs targeting genes in the PI3K ( Fig 5A and 5B ) and ERK ( S3 Fig ) pathways , the dual-layer integrated cell line-drug network model gave consistently better performance than CSN or DSN alone . In addition , for around half of these drugs , the Pearson correlations between the observed and predicted responses from the dual-layer integrated cell line-drug network model were higher than 0 . 5 ( Fig 5A and S3A Fig ) . Scatter plots of 4 example drugs also indicate that the good correlations are fairly reasonable with relatively small numbers of outliers ( Fig 5C and S3C Fig ) , results of the rest drugs in these two pathways are shown in ( S4 Fig ) . Similar conclusion can be drawn when using IC50 as drug response measurement ( S5 Fig ) . We next compared our network models with a machine-learning approach which also predicts drug response based on genomic feature of cell lines and chemical features of drugs [17] . Basically , it treats drug response prediction as a machine-learning problem where each possible drug-cell line pair is represented through integrating genomic features of the cell line and chemical structural features of the drug . Neural network ( NN ) and random forest ( RF ) are then used to build a prediction model based on training data . However , NN gave very poor performance despite our numerous attempts to set different parameters , so we test RF and another user-friendly machine-learning tool , support vector regression ( SVR ) . For CCLE and CGP datasets , the overall correlation between observed activity areas and predicted from SVR and RF ( S6 Fig ) among all drugs and all cell lines is high and almost the same as [17] . However , since the baseline of different drugs have very different activity area ranges ( S7 Fig ) , this overall correlation doesn’t indicate how a particular drug will behave on different cells . For each individual drug , the correlations are only around 0 . 20 ( RF ) and 0 . 23 ( SVR ) in CGP , and 0 . 49 ( RF ) and 0 . 38 ( SVR ) in CCLE , which are often worse than our single layer model as well as the elastic net model , and much less than our integrated model ( Figs 4A , 4B , 5A and 5B ) . In addition , the method in their model treats every possible drug-cell line pair as an individual case , so it is computationally expensive for larger number of drugs and cell lines , especially for CGP . But our model only relies on correlations between cell lines or drugs , thus is computationally efficient . However , as pointed by many researches , leave-one-out cross-validation in regression will possibly overestimate the accuracy of predictions on out-of-sample observations due to over fitting . To assess the potential bias in above leave-one-out cross-validation , we randomly shuffled response values of all cell lines to each drug , then repeated our predictions with the same parameter settings [25] . As is shown in ( S8 Fig ) , for all drugs tested in CCLE and CGP , we observed only very weak correlations between real and predicted values ( absolute correlations are less than 0 . 08 for more than 95% drugs ) , indicated that our model is not biased by the cross-validation procedure . Although the CGP study used a total of 707 cell lines and 139 drugs , only 653 cell lines had expression profiles and only 124 drugs had chemical information available . Out of the possible 653 × 124 cell line-drug combinations , only 76% have corresponding drug response data . With the cell similarity and drug similarity data , we could use our dual-layer integrated cell line-drug network model to predict the missing activity areas ( Fig 6 ) and IC50 ( S9 Fig ) , with a particular focus on three MEK inhibitors AZD6244 , RDEA119 , and PD-0325901 , where a large number of response values were missing . When grouping the cell lines based on their BRAF mutation profiles , we found that the BRAF-mutated cell lines were significantly more sensitive to MEK inhibitors ( Fig 6 ) . These predictions were consistent with those in cell lines where response data were available , and were in agreement with previously published studies . The above findings suggest that our dual-layer integrated cell line-drug network model can be used to optimize the design of cell line screens with new drugs by combining in silico predicted response values from existing screen results and the structure of the new drugs .
In this study , we developed a dual-layer integrated cell line-drug network model to predict the response of cancer cell lines to drug treatments for experimental activity areas in the CCLE and IC50 values in the CGP study . The novelty of our method lies in the incorporation of two distinct sub-networks for predicting drug responses: 1 ) a cell similarity network ( CSN ) , based on similarities in gene expression profiles between cell lines , and 2 ) a drug similarity network ( DSN ) , based on similarities in chemical structural features between drugs . This method applies a weighted algorithm to model the predictor . Using the data from the CCLE and CGP studies as benchmarks , leave-one-out cross-validation showed that our dual-layer integrated cell line-drug network model consistently outperformed the elastic net model , suggesting that our model effectively captures the interplay in the cell line-drug network . Our model also reasonably assigned much lower activity areas for BRAF mutant cell lines than for BRAF wild-type cell lines for three different MEK1/2-inhibitors , which is in good agreement with current experimental reports . Our network model can be applied to predict the response of a new cell line to existing already tested drugs or to predict the response of an existing cell line to new drugs , thus potentially saving the cost in a drug-cell line screening . Compared to the existing elastic net regression , our dual-network model is based on the entire dataset rather than a single drug , so interactions between different drugs are considered . Furthermore , our model only needs correlations between cell lines or drugs as input , thus is not seriously affected by the huge dimensionality of genome-wide gene expression and copy number variation features . Despite these encouraging results , our model suffers from the following limitations , which we hope to address in the future . First , construction of the CSN relied solely on gene expression profile data , and future work to integrate somatic mutation information , epigenetic status [26 , 27] , and pathways could potentially improve the performance of our network models . Furthermore , some other dynamic and direct features , such as time-course gene expression [28] or protein signaling [29] , might better illustrate a cell line and potentially give a more reliable CSN . Second , construction of the DSN used the 1-D and 2-D structural features of each drug , which might provide too many features for small molecule drugs , and yet ignored the 3-D structure features which may be important for certain drugs . Third , our dual-network model is a local weighted model based on responses of existing cell lines to the query drug or ( and ) the query cell line to other similar drugs by chemical structure . So it does not work without the above two types of information , such as the response prediction for a new drug to a new cell line , which is more useful in real application cases . With increasing data on cell-line drug responses becoming available over time , and extended network models to address the above limitations , we hope this network-based approach will have much better predictive power and potentially be used in drug combination exploration [30] . Finally , our local weighted model could possibly shrink the range of predicted drug responses and thus lead to a quite large RMSE . For example , if response of a drug d and cell line c is the largest among all training data , the linear combination of response values from their neighborhood will definitely be smaller than the truth . However , in practical clinical settings , one may care more about the relative order rather than the absolute values of drug response data due to the batch effect of different experiments . So most work in drug response prediction used correlation between true and predicted values as a measure of effectiveness [13 , 17] , which might be a better measure than RMSE . In fact , even the published original data from CCLE and CGP have different magnitudes in IC50 for their common drugs ( S7 Fig ) .
The cancer genomic and drug response data used in this work are available from the Cancer Cell Line Encyclopedia ( CCLE ) [13] and the Cancer Genome Project ( CGP ) [14] . The CCLE dataset consists of large-scale genomic data , including gene expression profiles , mutation status , and copy number variation for 1 , 036 human cancer cell lines , and eight-point dose-response curves for 24 chemical compounds across 504 cell lines . Gene expression profiles and drug sensitivity data ( measured by area under dose-response curves ) were downloaded from the CCLE website ( http://www . broadinstitute . org/ccle ) . Both drug sensitivity measurements and gene expression profile data are available for 491 common cancer cell lines . The CGP dataset provided raw gene expression profiles for 789 cell lines , which we downloaded from AffyExpress under accession number E-MTAB-783 . We also obtained drug sensitivity measurements ( activity area ) and cell line annotation data from the CGP website ( http://www . cancerrxgene . org/downloads ) , which included data for 139 drugs and 707 cancer cell lines . The CGP study provided both drug sensitivity measurements and gene expression profile data for 653 common cancer cell lines . We downloaded the chemical structural information for each drug from PubChem , a database of chemical molecules and their activities in different biological assays . It contains validated chemical depiction information for 19 million unique compounds contributed from over 70 depositing organizations . We downloaded raw chemical property profiles ( SDF files ) for 23 drugs in the CCLE study and 124 drugs in the CGP study from the PubChem website . The 1-D and 2-D chemical structural features of each drug were retrieved using the PaDEL software program ( v2 . 11 , downloaded from the project website http://padel . nus . edu . sg/software/padeldescriptor/ ) with default settings . In detail , the 1D descriptors consist of compositional or constitutional molecular properties , e . g . , atom count , bond count , and molecular weight . The 2D descriptors include different quantitative properties of the topology such as Kappa shape indices [31] , Randic [32] and Wiener indices [33] ( S1 Table ) . Our integrated network for predicting drug responses consists of three parts: 1 ) a cell line similarity network ( CSN ) , which connects all cell lines based on their Pearson correlation coefficients of gene expression profiles; 2 ) a drug similarity network ( DSN ) , where associations between two drugs are measured by the Pearson correlation between their respective 1-D and 2-D structural properties; and 3 ) a drug-cell line response network , which connects the above two networks using experimentally determined drug response values ( measured as activity area and IC50 in both CCLE and CGP studies ) . The CSNs generated using CCLE or CGP data are both complete graphs of all cell lines , where interactions between two cell lines are measured by the Pearson correlation coefficients of their respective gene expression profiles . Similarly , the DSNs are also complete graphs of all available drugs , connected with their pairwise correlation between their respective 1-D and 2-D chemical structural features . The intermediate layer between the CSN and the DSN , referred to as the drug-cell line response network , is a bipartite graph of all cell lines and drugs , labeled with the response values ( corresponding to activity area and IC50 values ) . Note that the intermediated drug response network is not a complete bipartite graph due to some missing values in two experiments , particularly in the CGP dataset . In order to predict the response of a new cell line C to a known drug D , similar to the LWR ( Locally weighted linear regression ) , we take advantage of data for all neighboring cell lines to cell line C in the cell line similarity network ( CSN ) . For drug D that is being investigated , neighboring cell lines are excluded from the analysis if drug sensitivity data is not available for them . Based on the hypothesis that cell lines with similar gene expression profiles will respond similarly to the same drug , we propose a linear weighted model to approximate the sensitivity of cell line C to drug D as follows: Sens^CSN ( D , C ) =∑Ci≠CSens ( D , Ci ) wC ( C , Ci ) ∑Ci≠CwC ( C , Ci ) ( 1 ) where Sens ( D , Ci ) is the sensitivity data of cell line Ci against drug D , wC ( C , Ci ) is the weight parameter , and Ci is the sample that is associated with cell line C in the CSN . According to our model assumption , the most closely related cell line in terms of its gene expression profile will contribute much more to the prediction of Sens ( D , C ) compared to other cell lines . Therefore , the weight parameter should be an increase function with their correlation to C . A fairly standard choice for the weights is: wC ( C , Ci ) =e−[1−ρC ( C , Ci ) ]22σ2 ( 2 ) in which ρC ( C , Ci ) is the gene expression correlation between cell line C and Ci , σ is the bandwidth parameter controlling how quickly the weight of a training example falls off with distance of its Ci from the query point C . Here , if ρC ( C , Ci ) is close to 1 , wC ( C , Ci ) will also be close to 1 , implying that this cell line has a high impact in the evaluation of Sens ( D , C ) . On the contrary , if ρC ( C , Ci ) is small ( e . g . , close to 0 ) , wC ( C , Ci ) will be relatively small too . In this case , the corresponding cell line will have a weak contribution to the determination of the unknown cell line . In the previous section , we developed a model to predict the response of an unknown cell line to an existing drug based on the training data . However , in some cases , we may need to predict the response of a known cell line to a new drug . Based on a model similar to that used in the previous section , we propose a linear weighted model for predicting the sensitivity of known cell line C to a new drug D based on the known sensitivities of cell line C to all neighboring drugs to drug D . We developed a similar linear weighted model to predict the response of a cell line to a new drug based on its neighboring drugs in the drug similarity network ( DSN ) as follows: Sens^DSN ( D , C ) =∑Dj≠DSens ( Dj , C ) wD ( D , Dj ) ∑Dj≠DwD ( D , Dj ) ( 3 ) where the weight wD ( Dj , C ) is determined by the correlation between the chemical structural feature of drug D and drug Dj in the drug similarity network: wD ( D , Dj ) =e−[1−ρD ( D , Dj ) ]22τ2 ( 4 ) where τ is the bandwidth parameter . In the above two sections , we developed single-layer predictive models for determining the response of a cell line to a drug based on the cell line similarity network ( CSN ) and the drug similarity network ( DSN ) , depending on whether drug and cell line information are available . However , the results from either of these single-layer models may be not satisfactory when only one type of information ( either cell line similarity or drug similarity ) is considered . To make full use of the integrated network , we propose a linear weighted model to combine Sens^CSN ( D , C ) and Sens^DAN ( D , C ) as follows: Sens^ ( D , C ) =λ∙Sens^DSN ( D , C ) + ( 1−λ ) ∙Sens^CSN ( D , C ) ( 5 ) where λ is the combination weight , which can be optimized through leave-one-out cross-validation . The integrated model will be dominated by DSN if λ is close to 1 , and it will be dominated by CSN if λ is close to 0 . Two individual models are complementary when 0 < λ < 1 . We used the leave-one-out cross-validation method to determine the parameters and validate our predictor . In detail , each possible drug-cell line pair is singled out as the test dataset to measure the consistency between predicted and observed drug response with the model trained from the remaining data . There are three free parameters to be determined in our model , i . e . , σ and τ which measure the decay rate when correlations of cell line expression or drug descriptors decrease , and λ which measures the relative contribution of each single layer . In order to ensure a unified model , two decay parameters σ and τ are fixed for all drugs and cell lines in CSN and DSN respectively , but λ is optimized by each individual drug , allowing different relative contributions of two lays for different drugs . Three parameters are optimized in the following order to make sure drug response can be predicted by each individual layer and the whole integrated network . For the cell line similarity network , the decay parameter σ is optimized by minimizing sum of squared prediction errors for all possible drug-cell line combinations using Formula ( 1 ) as prediction model . In detail , the overall error function ( or cost function ) is defined as: J ( σ ) =∑D , C ( Sens ( D , C ) −Sens^CSN ( D , C ) ) 2 ( 6 ) where Sens ( D , C ) is the observed sensitivity value of cell line C to drug D , and Sens^CSN ( D , C ) is the predicted one through cell line similarity models using all other drug-cell line interactions as the training set . The best parameter was obtained by minimizing the error function as follows: σ^=argminσJ ( σ ) . Similarly , decay parameter τ for the bottom drug similarity network was determined by τ^=argminτ∑D , C ( Sens ( D , C ) −Sens^DSN ( D , C ) ) 2 ( 7 ) Finally , the best λ for each drug is obtained through minimizing the sum of squared errors after combining predictions from both CSN and DSN , i . e . , λ^=argminλ∑C ( Sens ( D , C ) −Sens^ ( D , C ) ) 2 . ( 8 ) After selecting the parameters , the prediction performance of our model was evaluated using Pearson correlation coefficient , root mean squared error ( RMSE ) and normalized root mean squared error ( NRMSE ) between predicted and observed drug responses for each drug . RMSE is the square root of the mean squared error , RMSE ( D ) =∑c ( Sens ( D , C ) −Sens^ ( D , C ) ) 2n , ( 9 ) and NRMSE is got by dividing the range of drug responses to facilitate the comparison between different drugs or predictive models , NRMSE ( D ) =RMSE ( D ) maxCSens ( D , C ) −minCSens ( D , C ) . ( 10 ) A higher Pearson correlation coefficient and lower RMSE/NRMSE indicate a better prediction performance of a method for a given drug .
|
In this study , using the Cancer Cell Line Encyclopedia ( CCLE ) and Cancer Genome Project ( CGP ) studies as benchmark datasets , we explored the application of similarity information between cell lines and drugs in drug response prediction . We found that similar cell lines by gene expression profiles exhibit similar response to the same drug . Meanwhile , drugs with similar chemical structures also show similar inhibitory effects across different cell lines . Based on the above observations , we proposed a dual-layer network and local weighted model to predict drug response of a cell line using proximal information of the drug-cell line network . The only three parameters of our model are optimized by leave-one-out cross-validation for each drug . Two case studies of MAPK and ERK signal pathways on CCLE dataset proved that the predicted-to-observed correlations of our dual-layer network model is significantly better than the previous predictor using elastic net model . Interestingly , predictions based on drug similarity network ( DSN ) alone were much better than those based on cell line similarity network ( CSN ) alone for most drugs , implying that drug similarities are more informative for drug response prediction than cell line similarities . Our network model can be applied to predict the response of a new cell line to existing already tested drugs or to predict the response of an existing cell line to new drugs , thus potentially saving the cost in a drug-cell line screening .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Predicting Anticancer Drug Responses Using a Dual-Layer Integrated Cell Line-Drug Network Model
|
Genetic exchange by recombination , or reassortment of genomic segments , has been shown to be an important process in RNA virus evolution , resulting often in important phenotypic changes affecting host range and virulence . However , data from numerous systems indicate that reassortant or recombinant genotypes could be selected against in virus populations and suggest that there is coadaptation among viral genes . Little is known about the factors affecting the frequency of reassortants and recombinants along the virus life cycle . We have explored this issue by estimating the frequency of reassortant and recombinant genotypes in experimental populations of Cucumber mosaic virus derived from mixed infections with four different pairs of isolates that differed in about 12% of their nucleotide sequence . Genetic composition of progeny populations were analyzed at various steps of the virus life cycle during host colonization: infection of leaf cells , cell-to-cell movement within the inoculated leaf , encapsidation of progeny genomes , and systemic movement to upper noninoculated leaves . Results indicated that reassortant frequencies do not correspond to random expectations and that selection operates against reassortant genotypes . The intensity of selection , estimated through the use of log-linear models , increased as host colonization progressed . No recombinant was detected in any progeny . Hence , results showed the existence of constraints to genetic exchange linked to various steps of the virus life cycle , so that genotypes with heterologous gene combinations were less fit and disappeared from the population . These results contribute to explain the low frequency of recombinants and reassortants in natural populations of many viruses , in spite of high rates of genetic exchange . More generally , the present work supports the hypothesis of coadaptation of gene complexes within the viral genomes .
Genetic exchange is , with mutation , a primary source of genetic variation and plays an important role in virus evolution . Viruses possess mechanisms for genetic exchange that make their reproduction “just as sexual as that in eukaryotes” [1]: whenever different genetic variants replicate in the same cell , genetic exchange can occur by recombination of genome regions that are switched between nucleotide strands , or by reassortment of complete genome segments in viruses with segmented genomes . Genetic exchange results in novel genetic combinations that could have important phenotypic effects . It has been documented repeatedly that genetic exchange can result in dramatic changes in the properties of the viruses , and recombinant and reassortant genotypes have been associated often with host range expansion , with host switches , or with increased pathogenicity . An outstanding example is the reassortment between avian and human strains of influenza , resulting in novel viruses with pandemic potential , which is responsible for the most serious respiratory disease pandemic in humans; but other examples abound for both animal and plant viruses ( e . g . , [2–11] ) . In addition , genetic exchange may counterbalance the effect of deleterious mutation accumulation in virus populations [12] , as initially shown by the classical work on bacteriophage Ø6 , which showed that reassortment opposed the progress of mutational load in the virus populations [13 , 14] . In spite of its potential importance , rates of genetic exchange in viruses have been seldom analyzed ( e . g . , [15] ) , and little is known about how factors related to the virus life cycle or to the environment may affect the frequency of the resulting new genotypes in the virus population [16 , 17] . We have addressed the analysis of the factors that determine the frequency of reassortant and recombinant genotypes in virus populations , using Cucumber mosaic virus ( CMV ) as an experimental system . CMV ( genus Cucumovirus , family Bromoviridae ) is a plant virus with a messenger-sense , single-stranded , three-segmented RNA genome . Each genome segment is encapsidated separately in an isometric particle . RNA1 and RNA2 encode proteins 1a and 2a , respectively , which are part of the virus replicase . RNA2 also encodes protein 2b , in a second open reading frame ( ORF ) overlapping that for protein 2a , which is a suppressor of the post-transcriptional gene-silencing defense of the host plant . RNA3 has two ORFs separated by a noncoding intergenic region , the 5′-most of which encodes the 3a movement protein , needed for cell-to-cell movement of the virus in the infected host . The second ORF of RNA3 encodes the coat protein ( CP ) , which besides its structural function , is required for cell-to-cell and systemic movement and for vector transmission . CMV has a very broad host range , is transmitted in a nonpersistent manner by many species of aphids , and is found worldwide as the causal agent of economically important epidemics in many vegetable , fruit , and fodder crops ( see [18] for a review ) . CMV isolates have been classified into three subgroups ( named IA , IB , and II ) according to sequence similarity between their genomic RNA3 [19] . Viable reassortants and recombinants can be obtained between CMV isolates belonging to subgroups IA , IB , and II [18] , and on the basis of phylogenetic analyses it has been proposed that reassortment of genomic segments has played an important role in the evolution of CMV and has contributed to the high genetic diversity found among CMV strains [20] . Both reassortants and recombinants between CMV isolates belonging to different subgroups have been reported to occur in nature , but analyses of the genetic structure of CMV field populations have shown that reassortants and most recombinants were present at low frequency , and data indicated that they were at a selective disadvantage [21 , 22] . We have analyzed the frequency of reassortant and recombinant genotypes in plants double infected with CMV isolates from subgroups IA and IB . Analyses were done at various stages during colonization of the host plant by the virus , in order to dissect the role of different steps in the virus life cycle in the fate of the new genotypes . Results show that as plant colonization progresses , selection for particular gene combinations increases and the frequency distribution of the various possible genotypes departs more and more from random . Selection operates against genotypes with heterologous gene combinations resulting from genetic exchange between the parental strains , supporting the hypothesis of coadapted gene complexes in the virus genome .
Frequency distributions of parental , reassortant , and recombinant genotypes in LLH progenies of the four analyzed isolate combinations ( see Table S1 ) were homogeneous ( i . e . , did not differ at 95% level of confidence ) between combinations I , III , and IV , or II , III , and IV . Genotype frequencies pooled over the four isolate combination progenies are presented in Table 1 ( LLH pooled progeny ) , which also indicate the frequency of alleles A and B at loci i , j , and k1 . k2 . Because the relative proportion of alleles A and B at each loci in the mixed inoculum , measured in terms of infectivity , was 0 . 5:0 . 5 , the expected frequency of each genotype under the hypothesis of random reassortment can be calculated from the combinatorial probability of these allelic proportions at each of the three loci , assuming independence ( linkage equilibrium ) in the distribution of the three genomic segments: 0 . 5 × 0 . 5 × 0 . 5 = 0 . 125 . Each of the parental and the six possible reassortant genotypes occurred in LLH progenies , but with large differences in frequency , so that in no case the genotype frequency distribution did fit that expected from random ( 0 . 125 ) under the linkage equilibrium hypothesis ( p < 0 . 0001 for every progeny ) . Genotypes with allele A at loci k1 . k2 represented 0 . 78 of the pooled progeny: genotypes AAA . A , BAA . A , and BBA . A were the most frequent ones , the frequency of ABA . A and of genotypes with allele B at loci k1 . k2 was always lower and similar . Genotype frequency distributions were analyzed by fitting log-linear models , which express the overall logarithmic-deviation of observed genotype frequencies ( once fitted by the model ) from those expected under the null hypothesis ( see Materials and Methods ) . Parameters αi , βj , and γk in the model ( Table 2 ) represent the positive or negative log-deviations from the random expected genotype frequency ( 0 . 125 ) that are due to changes in frequency of alleles A or B at loci i ( αi ) , j ( βj ) , and k1 . k2 ( γk ) ( i , j , k1 , and k2 = A , B ) . Parameters αβij , αγik , βγjk , and αβγijk represent additional log-deviations due to associations among loci ( i . e . , due to linkage disequilibrium ) from the linkage equilibrium expectation under the observed allele frequencies . For example , in the case of LLH pooled progeny , γA and γB ( Table 2 ) indicate , respectively , increase ( γA > 0 ) and decrease ( γB < 0 ) factors of eγA and eγB in frequencies of allele A and B at loci k1 . k2 from their random expectation of 0 . 5 . The overall effect due to the change in A:B proportion at this loci is expressed by a unique parameter γk = 0 . 63511 , calculated as half the difference γA − γB , which indicates an increase ( γk > 0 ) factor of e2γk in frequency of allele A over frequency of allele B . Parameters of the model for the LLH pooled progeny ( Table 2 ) showed a significant increase in frequency of allele A relative to allele B at loci k1 . k2 ( p < 0 . 0001 ) and at locus j ( βj = 0 . 19149 , p = 0 . 0005 ) , while allele A decreased in frequency relative to allele B at locus i ( αi = −0 . 28039 , p < 0 . 0001 ) . Models also indicated significant ( p = 0 . 007 ) homologous association between locus j in RNA2 and loci k1 . k2 in RNA3 , resulting in increased genotype frequency of pairs iAA . A and iBB . B and decreased frequency of pairs iAB . B and iBA . A ( i = A , B ) by factors of eβγAA , eβγBB , eβγAB , and eβγBA ( Table 2 ) from their linkage equilibrium expectation given the observed allele frequencies . The overall effect of this association , βγjk = 0 . 18611 , calculated as the half mean of βγAA − βγAB and βγBB − βγBA , represents an increase factor ( βγjk > 0 ) of e2βγij in the frequency of homologous over heterologous allele combinations at loci j and k1 . k2 . A standard measure of linkage disequilibrium is Q , which is directly calculated from βγjk ( see Materials and Methods ) and takes the value 0 . 3559 . Another common standard measure of linkage disequilibrium is D′ , which in the case of those two loci was D′ = 0 . 2397 ( see also Materials and Methods ) . Significance of model parameters for each isolate combination is indicated in Table S1 . No recombinants at RNA3 were found in the progeny of any of the four combinations of IA and IB isolates ( Tables 1 and S1 ) . Frequency of genotype distributions of TIL , VIL , and TSL progenies for all four isolate combinations are detailed in Tables S2 , S3 , and S4 , respectively , and indicated in Table 1 for the corresponding TIL , VIL , and TSL pooled progenies . Homogeneity of frequency distribution of genotypes was found for combinations I and II in TIL , for combinations II and IV in VIL , and for all four isolate combinations in TSL ( p = 0 . 054 ) , being highly homogeneous for combinations I , II , and IV in TSL ( p = 0 . 978 ) . In all progenies from the systemic host , genotypes AAA . A , BAA . A , and BBA . A were the most frequent , genotypes with allele B at loci k1 . k2 being at low frequency ( TIL and VIL ) , or not detected at all ( TSL ) . Frequency of genotypes AAA . A and BAA . A together made 0 . 73 of the total population in TIL and progressively increased to 0 . 82 in VIL and to 0 . 95 in TSL ( Table 1 ) . Figure 2 shows the variation of all genotype frequencies in the pooled progenies from TIL to VIL and TSL , as compared to LLH . Analysis of genotype frequency distributions in TIL , VIL , and TSL pooled progenies by log-linear models showed a significant increase of A:B allele proportions from their random expectation hypothesis ( 0 . 5:0 . 5 ) at loci j and k1 . k2 in TIL and VIL , or at the three loci i , j , k1 . k2 in TSL , as shown by model parameters ( Table 2 ) . A significant homologous association between loci i and j was detected in the three pooled progenies: αβij = 0 . 17584 , Q = 0 . 3379 , D′ = 0 . 2627 ( p = 0 . 0053 ) for TIL; αβij = 0 . 37680 , Q = 0 . 6373 , D′ = 0 . 5672 ( p < 0 . 0001 ) for VIL; and αβij = 0 . 33817 , Q = 0 . 5891 , D′ = 0 . 2528 ( p = 0 . 0166 ) for TSL . Variation of genotype frequency distributions , as systemic colonization of the host progressed could also be analyzed by log-linear models , estimating the deviations of genotype frequencies in a progeny from the expected frequencies resulting from allelic frequencies observed in the previous step in the virus life cycle , assuming a linkage equilibrium distribution . For example , deviations in TIL pooled progeny distribution were estimated from linkage equilibrium distribution of allelic frequencies in the LLH pooled progeny , taking it as the null hypothesis in the model . Parameters of the model could be easily computed as the difference between that in TIL and that in LLH , as estimated in both cases from the random expectation ( Table 2 ) . It was found that deviations of TIL from LLH expectation were significant at the three loci , indicating the increase of the relative frequency of allele A: αi = 0 . 26391 , βj = 0 . 35781 , γk = 1 . 09918 ( p < 0 . 0001 ) . Deviations of VIL progeny from TIL expectation were only significant at locus j: βj = 0 . 35781 ( p = 0 . 0002 ) . TSL pooled progeny was analyzed from both TIL and VIL expectation . Deviations of TSL from both TIL and VIL expectation were significant at loci i and j [αi = 0 . 94929 , βj = 0 . 97155 ( p < 0 . 0001 ) for TIL; αi = 0 . 93603 , βj = 0 . 70643 ( p < 0 . 0001 ) for VIL] . No recombinants at RNA3 were detected in any TIL , VIL , or TSL progeny ( Tables 1 and S2–S4 ) .
Research during the last two decades has shown the important role of genetic exchange by reassortment of genomic segments , or by recombination , in RNA virus evolution [2 , 3 , 16 , 17 , 23] . However , although there are reports of virus populations at linkage equilibrium [24] , data from numerous systems indicate that the frequency of reassortant genotypes in natural populations of viruses with segmented genomes widely departs from random expectations , and , similarly , frequency of recombinants is lower than in experiments in the culture plate or in the greenhouse [22 , 25 , 26] . In fact , evidence suggests that selection against heterologous gene combinations occurs , perhaps due to coadaptation between the genes within a viral genome [21 , 27–30] . Little is known about the factors that may operate along the virus life cycle affecting the final frequency of reassortant and recombinant genotypes and its epidemiological impact in nature [17] . Here , we address these questions . The approach was to estimate the frequency of reassortants and of recombinants at RNA3 , in descents of double inoculations with CMV isolates belonging to subgroup IA and subgroup IB , which differ by about 12% in nucleotide sequence [18] . This level of genetic divergence should not hinder the possibility of genetic exchange , as CMV isolates in subgroups IA , IB , and II are known to produce viable reassortants and recombinants , which under experimental conditions are infectious and will accumulate to similar levels as the parental strains in single infection [18] . Also , a recent report has demonstrated an extremely high frequency of reassortants in natural populations of highly diverged ( up to 50% ) viruses within the Cystoviridae family [24] . The experiments were done with four pairs of CMV isolates , and analyses of progenies were planned to identify constraints to genetic exchange at four steps in the virus life cycle: ( i ) infection of leaf cells , ( ii ) colonization of the inoculated leaf , which involves infection of leaf cells and cell-to-cell movement , ( iii ) encapsidation of viral RNA , and ( iv ) colonization of upper , noninoculated leaves , which involves systemic movement through the phloem . Step ( i ) was analyzed in a local-lesion host , C . quinoa , through the estimation of genotype frequencies in the resulting local-lesion populations ( populations LLH ) , and steps ( ii–iv ) were analyzed in a systemic host , tobacco , in progeny populations represented by total RNA preparations from inoculated or systemically infected leaves ( TIL and TSL , respectively ) or by virion RNA from inoculated leaves ( VIL ) . Results from our experiments confirm that coinfection with IA and IB CMV isolates in the four combinations assayed resulted in the production of viable reassortants in the six possible combinations ( Table 1 ) . However , the frequency distributions of the eight parental and reassortant genotypes departed largely from random even at the very beginning of the infection process , as shown by data from LLH progenies ( Tables 1 and S1 ) . Different genotype frequencies could reflect differences in the relative ability of genomic segments carrying alleles A and B to infect cells , or alternatively , differences in the capacity to infect the host plant among the genotypes , i . e . , differences in their relative fitness . Local-lesion assays had shown that infectivity of the parental isolates in each combination was equivalent . Hence , results from LLH progenies indicate differences in fitness among genotypes for the formation of a single local lesion , i . e . , the establishment of a successful infection at an initial cell , plus restricted virus movement to a few surrounding cells . Deviations from the null hypothesis could also be due to sampling bias during the isolation-infection processes , e . g . , selection against particular genotypes during amplification of single-lesion descents in tobacco , but this possibility most probably can be discarded , given the high number of lesions transferred and the high success of infection in tobacco from necrotic local lesions ( about 80% ) . Also , random genetic drift associated with sampling or with population bottlenecks during local-lesion cloning cannot be discarded , but , again , the high number of lesions transferred for each progeny should minimize the impact of drift in the obtained results . Hence , as far as departures from expectation can be attributed to selection , the quantity ah ( ijk1 . k2 ) = −1 + eϕ ( see Materials and Methods ) represents a coefficient of selection for each genotype , which can be calculated from parameters of the log-linear models . Differences in coefficients of selection among the different genotypes were even higher in the systemic host than in the local-lesion host , and increased as the infection process progressed , i . e . , in progeny populations TIL , VIL and TSL . Biological cloning of descents in C . quinoa should introduce a bias in genotype frequencies , as not all genotypes were equally infectious to this host , but the effect would be the same for populations TIL , VIL , and TSL . Thus , each step during colonization of the systemic host resulted in stronger genotype selection , so that genotype frequency distribution departed more and more from random , and only three genotypes were detected in the four TSL progenies . It is to be noted that differences of genotype frequency distribution among the progenies of the four analyzed isolate combinations occurred in TIL and VIL , but frequency distribution was homogeneous for the four TSL progenies , indicating that similar selective pressures were operating in each progeny regardless of the nature of the parental isolates . Nonrandom distribution of reassortants in experimental populations has been reported for other RNA viruses with segmented genomes and has been interpreted as due to specific associations between genomic segments related to functional interactions between the RNAs or their protein products [27 , 30 , 31] . Alternatively , nonrandom reassortment has been explained as due to selective advantages of specific genome segments [29 , 32–34] . Because differences in selective advantage between genomic segments will not be independent of genetic context , both interpretations share a common basis and relate to the concept of coadaptation of gene complexes within the viral genome [35] . Occurrence of epistasis and coadapted gene complexes in genomes has important consequences for the evolution of natural populations [35] , and efforts have been made to estimate epistasis on viral genomes , mostly based on analysis of fitness effects of two or more point mutations [36 , 37] , rather than on the fitness of hybrids in progenies from crosses . Our data indicate an advantage of allele A over allele B at loci k1 . k2 in RNA3 since the earlier stages of infection ( populations LLH and TIL , Table 1 ) , so that in systemically infected leaves ( TSL populations ) genotypes with allele B at loci k1 . k2 were not detected ( Table 1 ) . This was also the case for locus j in RNA2 , the advantage of allele A , since the earlier stages of infection were particularly noticeable after encapsidation . Only at locus i in RNA 1 was allele B not at disadvantage relative to allele A in the inoculated leaves , but allele B was at disadvantage in systemically infected leaves , where the parental genotype AAA . A prevailed in the four progenies ( Table 1 ) . Selection against particular alleles at the analyzed loci was not independent of their genetic background , as significant associations of homologous gene combinations , indicating linkage disequilibrium , were found for some loci ( j and k1 . k2 or i and j ) in all pooled progenies . Hence , our data support the hypothesis of coadaptation of the four analyzed genes within the CMV genome , and show the higher fitness of genotype AAA . A in the assayed host and conditions . This result is in agreement with a recent report on the diminished competitive ability of a CMV reassortant [38] . The results of the present work also agree with analyses of the genetic structure of field populations of CMV , where isolates from subgroups IA and IB were at similar frequencies and were often coinfecting the same plant , but selection against reassortants and most recombinants ( i . e . , against heterologous gene combinations ) seemed to occur [21 , 22] . Mechanisms for nonrandom association of genomic segments have been proposed for other viruses , with selective advantages for specific genome segments being functionally linked to differences in replication efficiency [32 , 33] , infectivity after assembly into particles [34] , or interaction with host cell factors [29] . For CMV , allele A at loci k1 . k2 in RNA3 could have an advantage relative to allele B in competition for infection sites , replication , or cell-to-cell movement . RNA3 encodes promoters and other regulatory sequences for its replication as well as the two proteins , 3a and CP , required for cell-to-cell movement [18] . The relative increase in frequency of allele A relative to allele B at locus j in RNA2 as cell-to-cell movement progresses ( compare data for LLH and TIL ) or during encapsidation ( compare data for TIL and VIL ) , suggests that the homologous combination of RNA2 and RNA3 performs better than the heterologous cell-to-cell movement and encapsidation functions . This would not be the case for the combination of RNA1 and RNA3 . There was a sharp increase in relative frequency of allele A at locus i in RNA1 associated to systemic movement , suggesting that the homologous combination of the three RNAs performs the function of systemic colonization better , which occurs in the form of assembled viral particles and may depend on interactions between the CMV capsid and host factors [39 , 40] . Alternatively , a higher fitness for homologous allele combinations in RNAs 1 and 2 , related to the interaction of their protein products in the viral replicase [41] , could lead to a delayed increase of allele A at locus i as it increases at locus j , and explain the association between loci i and j observed in the systemic host . We have also analyzed the frequency of RNA3 recombinants between loci k1 and k2 in the progenies of the four IA and IB CMV isolate combinations . No recombinant genotype was detected among 1 , 381 descendents in LLH , TIL , VIL , or TSL progenies . Thus , the probability of finding one recombinant RNA3 was lower than 0 . 004 , at a 95% confidence level , assuming that this probability was the same for the 16 progenies . Otherwise , the probability would be even lower and the 0 . 004 value would be an upper threshold estimate . The observed differences in fitness between type A and B RNA3 , since the earlier stages of infection , might affect the probability of coinfection in the same cell , and , thus , the probability of recombination , as reported for CMV and Tomato aspermy virus [42 , 43] . In addition , exclusion of different CMV strains from infected cells [44] would also decrease the probability of recombination . In the progeny populations analyzed in this work , heterologous allele combinations were underrepresented relative to their expectation under the null hypothesis of linkage equilibrium . This shows the existence of constraints to genetic exchange linked to the various steps of host infections and colonization . Our results show that , whatever the reason for an initial disadvantage of genotypes with heterologous gene combinations , selection against these genotypes becomes stronger as host colonization proceeds so that the fittest genotypes would be the most available for host-to-host transmission and the heterologous gene combinations would disappear from the population . These results are important for understanding the role of genetic exchange in virus evolution and may be relevant for applied aspects of plant virology , as they might affect the durability of resistance genes [45] , or the ecological risks of virus-resistant transgenic plants [46] . In a more general context , these results support the hypothesis of coadaptation of gene complexes within a genome , which might be particularly relevant for the small , compacted , nonredundant genomes of RNA viruses .
Eight CMV isolates were used in this work , four belonging to subgroup IA and four to subgroup IB . These isolates were derived from field-infected zucchini squash or tomato plants sampled in Spain between 1992 and 1994 , when both types of isolates were frequent in the field [21] and were characterized as belonging to subgroups IA or IB by ribonuclease protection assay as described [21] . Isolates were multiplied in Nicotiana tabacum cv . Xanthi-nc , virion stocks were purified from systemically infected leaves as in [47] , and virion RNA was extracted with phenol and sodium dodecyl sulfate . Genetic exchange was analyzed in progenies from double inoculations with four different pair combinations of one isolate belonging to subgroup IA and one isolate belonging to subgroup IB . Coinoculations were replicated in ten half-leaves of the local-lesion host Chenopodium quinoa , and in five plants of the systemic host Nicotiana tabacum cv . Xanthi-nc . All inoculations were with virion RNA in 0 . 1 M Na2HPO4 , in leaves previously dusted with carborundum . RNA concentration of each isolate in the inoculum was such that the isolate's infectivity ratio was 0 . 5:0 . 5 . The relative infectivity of the IA and IB isolates in each combination was estimated by local-lesion assays in ten half-leaves at three different inoculum concentrations in the range 0 . 1–2 . 5 μg RNA/ml . For the tobacco plants , total RNA and virion-encapsidated RNA was purified from inoculated leaves 7 d post inoculation ( dpi ) , and total RNA was purified from systemically infected leaves 12 dpi . Total RNA was extracted from 200 mg of plant tissue as in [48] . Virus particles and virion RNA were purified as in [49] from inoculated leaves . In this way , three different RNA preparations were obtained from each infected plant , representing different progeny populations: total RNA from inoculated leaves ( TIL ) , virion RNA from inoculated leaves ( VIL ) , and total RNA from systemically infected leaves ( TSL ) . Total or virion RNA from each of the five infected plants per treatment was pooled , diluted at a ratio of 50 mg tissue/ml , and inoculated onto half-leaves of C . quinoa , for the cloning of single-lesion descendents . About 150 local lesions per progeny were individually transferred to small Xanthi-nc tobacco plants for multiplication , and 15 d later total RNA was extracted from these tobacco plants for the genetic characterization of descendents . Four pairs of oligonucleotide probes specific for ORFs encoding proteins 1a , 2a , 3a , and CP of CMV isolates in subgroups IA and IB were designed on the basis of nucleotide sequence information from ten CMV isolates of subgroup IA and eight CMV isolates of subgroup IB [21 , 50] ( unpublished data ) : the first pair , CMV1A ( 5′CATTAATGTCTATTCG3′ ) and CMV1B ( 5′CGTTGATGTCGATACG3′ ) were complementary to positions 1 , 330–1 , 346 of CMV RNA1; the second pair , CMV2A ( 5′GCGCTGTGAATAACGG3′ ) and CMV2B ( 5′GCGCAGTAAACAACGG3′ ) were complementary to positions 1 , 506–1 , 521 of CMV RNA2; the third pair , CMV3a-A ( 5′GACCCTTCAGCATCAG3′ ) and CMV3a-B ( 5′GATCCCTCAGCGTCGG3′ ) were complementary to positions 421–436 of CMV RNA3; the fourth pair , CMVCP-A ( 5′GGACTCCAGATGCGGC3′ ) and CMVCP-B ( 5′GAACGCCGGATGCAGC3′ ) were complementary to positions 1 , 722–1 , 737 of CMV RNA3 . Dot blot hybridization with these eight oligonucleotide probes , 5′-labeled with 32P [51] , unequivocally identified genetic types IA and IB in the four analyzed ORFs for the eight IA and IB parental CMV isolates ( unpublished data ) . Frequency distributions of genotypes in progenies were compared to expected frequency distributions according to the null hypothesis being tested , which was derived from the frequencies of IA and IB genetic types at each analyzed CMV ORF: either 0 . 5:0 . 5 at the inoculum , or as resulted from genotype distributions in previous steps of the virus life cycle , always assuming linkage equilibrium . Comparison between observed and expected frequencies was performed by the chi-square ( χ2 ) goodness of fit test . Comparison of frequency distributions of genotypes for the four CMV isolate combinations was done by the log-likelihood ratio test ( G ) for homogeneity of replicates tested for goodness of fit [52] . Genotype distributions were analyzed for independence among genomic segments ( linkage equilibrium ) by three-way contingency tables , which were solved upon the use of log-linear models [52] . These models were adapted to take the form: Ln[F ( ijk1 . k2 ) ] = Ln[Fh ( ijk1 . k2 ) ] + ϕh , where F ( ijk1 . k2 ) is the model estimate for the observed frequency of genotype ijk1 . k2 ( i , j , k1 , and k2 are loci 1a , 2a , 3a , and CP , respectively , and may take the allelic values A for genetic type IA and B for genetic type IB ) ; Fh ( ijk1 . k2 ) is the expected frequency of that genotype under the null hypothesis h , and ϕh is the overall log-frequency deviation under that hypothesis , where ϕh = αi + βj + γk + αβij + αγik + βγjk + αβγijk ; αi , βj , γk are log-deviations due to the frequencies of alleles A or B at loci i , j , and k1 . k2 and αβij , αγik , βγjk , and αβγijk are the log-deviations due to associations among loci in different genomic segments: i , j , and k1 . k2 . Deviation parameters were computed by fitting the model to observed and expected frequencies for each genotype [52] . When significant association among genomic segments was found , linkage disequilibrium was measured by the two standard metrics Q and D′ [53]: Q may be directly computed from association parameters in the model as Q = ( λ−1 ) / ( λ+1 ) , where λ = e4·A , A being the model parameter; D′ measures linkage disequilibrium relative to its maximum value under the observed allele frequencies , D′ = D/Dmax ( for a set of two biallelic loci in which alleles A and B have frequencies pA+ and pB+ at the first locus , p+A and p+B at the second locus , and pAA , pAB , pBA , pBB are frequencies of the four possible genotype combinations , then D = pAA pBB – pAB pBA , and Dmax is the lesser of pA+ p+B and p+A pB+ if D is positive , or the lesser of pA+ p+A and pB+ p+B if D is negative [53] ) . The quotient F ( ijk1 . k2 ) /Fh ( ijk1 . k2 ) = 1 + ah ( ijk1 . k2 ) represents the relative departure from expectation hypothesis h . Its deviation from one would be a coefficient of selection in case of fitness variation , which can be computed as ah ( ijk1 . k2 ) = −1 + eϕ . For all statistical tests , the probability of rejecting the null hypothesis was calculated by χ2 or G exact methods , or by Monte Carlo simulations with 106 replicates , using the SAS Statistical v 9 . 1 package ( SAS Institute , http://www . sas . com ) .
The GenBank ( http://www . ncbi . nlm . nih . gov/Genbank ) accession numbers for the nucleotide positions in the sequences for CMV discussed in this paper are RNA1 ( D00356 ) , RNA2 ( D00355 ) , and RNA3 ( D10538 ) .
|
The exchange of genomic regions between viral strains or species is an important process in virus evolution , resulting often in dramatic changes in virulence and host range and in the emergence of new viral diseases . In spite of its potential importance , little is known about what factors affect the frequency of the resulting new genotypes in virus populations . We explore this issue using Cucumber mosaic virus , a plant virus with a tripartite RNA genome . Experimental populations were derived from mixed infections with different strains and were analyzed at different moments during host colonization . Results showed the existence of constraints to genetic exchange linked to various steps of the virus life cycle . These results contribute to explain the often low frequency of recombinant and reassortant genotypes in natural populations of many viruses , in spite of high rates of genetic exchange , and support the hypothesis of coadaptation of gene complexes within the viral genomes .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] |
[
"viruses",
"virology",
"evolutionary",
"biology",
"plants",
"genetics",
"and",
"genomics"
] |
2007
|
Constraints to Genetic Exchange Support Gene Coadaptation in a Tripartite RNA Virus
|
Sleeping sickness is spread over 36 Sub-Saharan African countries . In West and Central Africa , the disease is caused by Trypanosoma brucei gambiense , which produces a chronic clinical manifestation . The Luba focus ( Bioko Island , Equatorial Guinea ) has not reported autochthonous sleeping sickness cases since 1995 , but given the complexity of the epidemiological cycle , the elimination of the parasite in the environment is difficult to categorically ensure . The aim of this work is to assess , by a molecular approach ( Polymerase Chain Reaction , PCR ) , the possible permanence of T . b . gambiense in the vector ( Glossina spp . ) and domestic fauna in order to improve our understanding of the epidemiological situation of the disease in an isolated focus considered to be under control . The results obtained show the absence of the parasite in peridomestic livestock but its presence , although at very low rate , in the vector . On the other hand , interesting entomological data highlight that an elevated concentration of tsetse flies was observed in two out of the ten villages considered to be in the focus . These findings demonstrate that even in conditions of apparent control , a complete parasite clearance is difficult to achieve . Further investigations must be focused on animal reservoirs which could allow the parasites to persist without leading to human cases . In Luba , where domestic livestock are scarcer than other foci in mainland Equatorial Guinea , the epidemiological significance of wild fauna should be assessed to establish their role in the maintenance of the infection .
Human African Trypanosomiasis ( HAT ) , also known as sleeping sickness , is a parasitic disease endemic of the African continent . HAT is caused by two subspecies of the flagellate Trypanosoma brucei; T . b . gambiense , spread over West and Central Africa , which is responsible for the chronic form of the disease ( more than 90% of total number of cases ) and T . b . rhodesiense , which is present in East Africa and produces a few cases of acute infection per year . In addition , other members of Trypanosoma genus are able to infect a wide variety of animals producing diseases of veterinary importance such as nagana ( T . b . brucei , T . vivax and T . congolense ) , surra ( T . evansi ) or dourine ( T . equiperdum ) . T . brucei s . l . is mainly transmitted by tsetse flies ( Diptera , Glossinidae ) but other trypanosomes can be mechanically or sexually transmitted [1] , [2] . In last years , control activities against sleeping sickness have been encouraged and significant advances were achieved to eliminate the disease [3] . The main strategy was to actively screen the human carriers in endemic foci [4] , [5] since it is assumed that humans are the main reservoir of T . b . gambiense infection [6] . Luba focus , located on Bioko Island ( Equatorial Guinea ) , is a good example of the success of control campaigns exclusively directed to humans . HAT was firstly declared in Luba in 1910 [7] and two decades later a successful control programme was implemented . At the end of 1960s , sleeping sickness was considered to be under control over the entire country and after the independence in 1968 , HAT ceased to be a priority of public health for the new authority . In the middle of 1980s , Luba suffered a resurgence of the disease registering hundreds of cases leading to the establishment of the Sleeping Sickness National Control Programme ( SSNCP ) by Health Ministry and Social Welfare . This programme targeted the disease control combining active cases detection and passive surveillance using serological techniques . All parasitologically confirmed and serologically suspected cases were treated . This strategy led to a drastic reduction in the number of reported patients . The last autochthonous case was recorded in 1995 and no more surveys have been carried out since 2004 [5] . As occurred in Luba , the neglect of control activities in the past has led to a resurgence of HAT foci considered to have been eliminated . Several hypotheses could explain the resurgence of the disease in apparently controlled foci and the heterogeneity of the disease prevalence in neighbouring foci: movement of carrier populations from active foci [8] , changes of the tsetse flies host preference [9] , [10] , genetic variability of the parasite [11] , [12] , the existence of asymptomatic parasite-infected individuals [13] , inherent limitations of surveillance systems [14] and maintenance of infection in animal reservoirs . The latter theory is supported by the capability of the parasite for surviving in some species of domestic and wild animals [15]–[24] . This study aims to analyse T . b . gambiense infection in tsetse flies and domestic livestock from localities of Luba focus , in order to determine the presence of the parasites apart from the human transmission cycle . Species-specific molecular tools ( PCR ) were employed for diagnosis . In addition , entomological data about tsetse fly populations in these localities are provided and discussed .
Luba focus covers a surface of 700 Km2 in south-western of Bioko Island . There are two climatic seasons: the dry season , from December to May , and the rainy season , from June to November . Bioko's annual rainfall exceeds 2 , 000 mm and the relative humidity ranges from 70% to 100% throughout the year . The average temperature is 25°C , with the minimum ranging from 17°C to 21°C and the maximum from 29°C to 30°C , depending on the location and the season [25] . The majority of the inhabitants from Luba lived on smallholding , hunting and sea fishing . Rainforest and neglected cocoa plantations are widespread in Luba district , establishing suitable habitats for the tsetse flies [26] , [27] . Nowadays , many people from rural areas have migrated to urban , mainly to the capital city ( Malabo ) , due to the recent development of petroleum and building industry . Therefore , rural conditions have partially disappeared and , as a result , many risk factors have been removed . Nevertheless , some villages remain unchanged and human-vector contact is still common . In September 2007 , blood samples of domestic animals ( pigs , sheep and goats ) were collected and tsetse flies were captured for further molecular analysis . Sampling procedures on Whatman filter paper for animal blood have been described elsewhere [28] . A previous census of livestock was elaborated in order to ensure a significant sample size . Ethical approval was obtained by Ministry of Health and Social Welfare and Veterinary Service from continental region ( Ministry of Agriculture , Forestry and Environment ) . The study was conducted adhering to these institutions' guidelines for animal husbandry . Verbal informed consent was obtained from each owner of livestock prior to the extraction of blood samples by the field team . Monopyramidal traps were employed to catch tsetse flies [29] . This kind of trap has been successfully applied for vector control and entomological surveillance in Equatorial Guinea [26] , [30] , [31] . This device makes flies to fall in a collecting bottle containing conservation solution ( formaldehyde 5% ) and to be stored until the gathering . Fifty-five traps were spread over the 10 villages belonging to epicentre of Luba focus . They were located in places considered a priori as suitable habitats for tsetse flies . This criterion includes water sources , cocoa or coffee plantations and shady and humid ponds close to livestock [32]–[34] . Geographical coordinates of all traps were registered by GPS ( Figure 1 ) and they remained two weeks in the field . Sampling was carried out twice , a week after the placing and when the traps were removed . Tsetse flies collected were stored in tubes with absolute ethanol in the field and separately processed in laboratory recording the trap number , an individual code , village , date , species , sex and age . The key of Brunhes et al . [35] was used for species identification and an age estimator , based on the degree of wear or fraying observed on the hind margin of the wing , was employed as previously described [36] , [37] . In addition , apparent density ( AD ) , estimated as AD = number of tsetse flies/trap/day , was calculated for each trap . Both tsetse flies and blood samples were sent to National Centre of Tropical Medicine , Institute of Health Carlos III ( Spain ) for molecular processing . DNA extraction from blood samples was performed employing a slightly modified protocol with Chelex 100® ionic resin ( Bio-Rad Laboratories , Madrid , Spain ) as previously described [28] , [38] . Prior to DNA extraction , wings and legs of tsetse flies were removed using a sterile surgical blade . This step was carried out in order to minimize the amount of exoskeleton compounds included in the sample , which are known to inhibit subsequent enzymatic reactions [39] . Flies were then washed in 70% ethanol and in double distilled water ( DDW ) . For dried samples , each fly was put in a sterile 1 . 5 ml tube and DNA extraction was performed employing the SpeedTools Tissue DNA Kit ( Biotools , B & M Labs , S . A . , Madrid , Spain ) following the manufacturer instructions . All extraction instruments were sterilized after processing each fly by ethanol submersion and flaming . Finally , a negative control ( clean 1 . 5 ml tube with no sample ) was included in all procedures of the extraction ( one negative each seventeen samples ) . Ten µl of DNA template from blood samples were subjected to species-specific PCR for T . brucei s . l . and , when positive , for T . b . gambiense . For T . brucei s . l . analysis , TBR1/2 primers were used [40] with the following conditions: 1× PCR Reaction Buffer ( 10 mM Tris-HCl , 1 . 5 mM MgCl2 , 50 mM KCl , pH 8 . 3 ) , 200 µM of each deoxynucleotide ( dNTP ) , primers at 0 . 5 µM and 1 . 25 U of Taq DNA Polymerase ( Roche Diagnostics , S . L . Barcelona , Spain ) in a final volume of 50 µl . Positive samples for this test were diagnosed for T . b . gambiense employing a nested-PCR with a first reaction using TgsGP1/2 primers [41] and a second one with TgsGP sense2/antisense2 primers described by Morrison et al . [42] . In both reactions 50 µl of final volume were reached and conditions were identical to T . brucei s . l . test with the exception of the amount of polymerase employed ( 2 . 5 U ) . The amplification programme for T . brucei s . l . was set as follows: a first step at 85°C ( 5 min ) for hot starting , 3 min at 95°C for initial DNA denaturation , 40 cycles of 95°C ( 1 min ) , 55°C ( 1 min ) and 72°C ( 1 min ) and a final extension step at 72°C ( 5 min ) . For the first reaction of T . b . gambiense the fixed programme was: initial denaturation step at 95°C ( 5 min ) , 45 cycles of 94°C ( 1 min ) , 63°C ( 1 min ) and 72°C ( 1 min ) with a final extension step at 72°C for 5 min . The programme for the second reaction was identical but only 25 cycles were performed . The quality of DNA templates from tsetse samples were tested by amplification of specific tubulin gene following the protocols described by Hao Z et al . ( 2003 ) [43] and Ferreira F et al . ( 2008 ) [44] . This step was considered because of the known PCR inhibition with samples of arthropods [39] . DNA samples that displayed a positive amplification signal for the tsetse tubulin gene were further tested to detect T . brucei s . l . and T . b . gambiense with the same primers and similar conditions as above: 1× PCR reaction buffer ( 10 mM Tris-HCl , 50 mM KCl , pH 8 . 3 ) , 2 mM MgCl2 , primers at 0 . 5 µM , 200 µM of each dNTP , 1 µl of DNA template and 1 U of AmpliTaq® Gold DNA Polymerase ( Applied Biosystems , Branchburg , New Jersey , USA ) reaching a final volume of 25 µl . The amplification programmes were modified increasing the time for the first denaturation step up to 10 minutes for polymerase activation as recommended by manufacturer and excluding the manual hot start step performed previously for T . brucei s . l . reactions . In all PCR assays two negative controls ( with DDW as template ) and one positive control ( 1 ng DNA from T . brucei s . l . 328 . 114 or 1 ng from T . b . gambiense ELIANE ) were included . All amplification products were separated by electrophoresis in a 2% agarose gel stained with ethidium bromide ( 1 µg/ml ) and photographed under UV light . Any sample displaying a visible band of the expected length in the gel was considered as positive ( Figure 2 ) . SPSS software ( version 16 . 0 . 1 , SPSS Inc . , Chicago , IL , USA ) was employed for statistical analysis and randomization of samples . Chi-square analysis was applied in order to compare significance of differences between variables ( CI 95% ) . Given the low number of animals sampled in each village and subsequently positives for T . brucei s . l . , no variables were statistically compared . Regarding tsetse flies , statistical associations between sex , age , sampling week , village and T . brucei infection were analysed , considering statistically significant p-value<0 . 05 .
Overall 1 , 839 flies were collected . Almost all of them ( 1 , 830 ) were identified as G . palpalis palpalis as expected by previous entomological data in the country [5] , [30] , [31] . Eight flies were registered as Glossina caliginea and one individual could not be classified . Only two villages ( Bococo Drumen and Bococo Avendaño ) concentrated 89 . 7% of overall captures ( 1 , 650/1 , 839 ) ( Table 1 ) and those localities showed a mean AD higher than 1 ( 14 . 9 in B . Drumen and 8 . 7 in B . Avendaño ) . The overall sex ratio was 1 . 59 ( 1 , 128 females/711 males ) and no significant difference was observed between the sampling weeks ( χ2 = 0 . 005 , p = 0 . 942 ) . However , sex ratio varies among the villages , from 1 . 01 in the lowest sampling localities to 1 . 33 in B . Avendaño and 2 . 2 in B . Drumen ( χ2 = 19 . 2 , p<0 . 001 ) . Distribution of flies by age groups shows that during second sampling week collected individuals were significantly younger than in the first one ( χ2 = 62 . 16 , p<0 . 001 ) . A total of 951 tsetse flies gathered ( 761 randomly selected from B . Drumen and B . Avendaño and 190 flies collected from the other villages ) were submitted to DNA extraction and specific tubulin amplification . From them , 905 ( 95 . 2% ) yielded a positive result for tubulin amplification and then were considered to have enough DNA quality to be included in further diagnostic analysis . Overall , 28 . 6% ( 259 ) of these flies were positive for specific T . brucei s . l . PCR and only one ( a young male fly from Bococo Drumen ) showed a positive amplification for TgsGP being considered as carrier of T . b . gambiense . No significant difference in T . brucei s . l . prevalence was observed regarding the sex ( χ2 = 0 . 708 , p = 0 . 401 ) , age ( χ2 = 9 . 368 , p = 0 . 095 ) or week of sampling ( χ2 = 0 . 000 , p = 1 . 00 ) but infection rate was significantly higher in B . Drumen than the others localities ( χ2 = 42 . 43 , p<0 . 001 ) . Among the G . caliginea individuals ( 6 ) included in the analysis neither T . brucei s . l . nor T . b . gambiense were detected . A previous census showed that there were 161 animals in villages belonging to the epicentre of Luba focus . Only 84 ( 52 . 2% ) could be sampled since livestock were not kept in sheds and moved freely around the dwellings . Nine animals ( eight pigs and one goat ) ( 10 . 7% ) yielded a positive result for T . brucei s . l . and none of them resulted positive for T . b . gambiense ( Table 2 ) .
In order to investigate the presence of Trypanosoma brucei gambiense in Luba focus and its possible maintenance in a non-human transmission cycle , molecular diagnosis was performed on tsetse flies and peridomestic fauna . Our study has revealed the occurrence of one tsetse fly carrier of T . b . gambiense , demonstrating that parasite has not completely disappeared from the environment . Although only one individual was considered positive ( prevalence ∼0 . 1% ) this result is consistent with the typical low infection rate reached in the vector even in active foci [45] , [46] . Regarding the domestic livestock no T . b . gambiense was detected , contrary to occur in other continental foci [28] . Prevalence of T . brucei s . l . was also determined in order to obtain an estimation of transmission activity in the focus . In tsetse flies , an overall high rate of infection was shown , although variations were observed regarding the village . A significantly higher tsetse flies infection rate was observed in Bococo Drumen from where it was gathered the majority of specimens collected in the focus during the study . In contrast , a relatively low prevalence of T . brucei s . l . in livestock was noticed , especially when compared with those previously described in continental foci [28] . Taking both data together it could be suggested that there is a high transmission activity , mainly in Bococo Drumen , but domestic fauna do not seem to act as the main feeding source for G . p . palpalis populations as in the other mainland foci . A wild transmission cycle could explain all these epidemiological features: G . p . palpalis would mainly feed over wild fauna , which could lead to a high trypanosome infection rate of the vector and the maintenance of T . b . gambiense in the focus . Wild fauna would have a role as reservoir of T . b . gambiense , absent in peridomestic cycle , and it would be expected to show a higher prevalence for T . brucei s . l . than observed in livestock . In contrast with this situation , in Mbini focus ( mainland Equatorial Guinea ) domestic animals ( sheeps and goats ) have shown to be carriers of T . b . gambiense [28] . It should be pointed out that livestock breeding is more common in continental foci , where almost all villages have some kind of farming . In Mbini , around five hundred animals were censed in a previous study [28] , whereas Luba only registered 161 . A less availability of livestock could be other factor which favours the feeding preference of G . p . palpalis for wild fauna . Tsetse fly host preferences should be thoroughly studied in order to clarify this issue but the opportunistic feeding behaviour of this species described in several studies allows to hypothesize about a wild transmission cycle [47] , [48] . Also , further research should be carried out in Luba to determine the role of wild animals in the maintenance of T . b . gambiense which has been previously described and discussed in neighbouring countries [15]–[17] , [20] , [22]–[24] . Some points about data must be added to this discussion . Firstly , the relative low sensitivity of the diagnostic technique employed for T . b . gambiense detection was previously reported since it targets TgsGP gene , only present once per haploid genome [49] . The low sensitivity of the test should not be a priori a serious drawback for tsetse flies samples since a great number of Trypanozoon subgenus parasites are usually found in infected midgut ( around a maximum of 106 ) [50] but it should be taken into account when data from animal blood samples are analysed . On the other hand , not all domestic animals were sampled and , as a result , it cannot be categorically ruled out the presence of the parasite in these hosts . These considerations should lead to weigh up the alternative hypothesis of the maintenance of infection in domestic livestock even in the absent of positive samples . In favour of this theory , it is noteworthy that Bococo Drumen and Bococo Avendaño , the villages where the tsetse fly density was higher , show the majority of domestic animals too . However , the low T . brucei s . l . infection rate found in livestock contradicts this apparent correlation . Future studies narrowly focused in feeding preferences of tsetse flies in Luba could clarify this issue . Entomological data were also gathered in order to better understand the epidemiological dynamic of the focus . Surprisingly , it was shown that only two out of ten villages sampled in the epicentre of Luba exhibited a high density of vector population . Almost 90% of collected flies belonged to B . Drumen and B . Avendaño . These villages registered a significantly higher AD levels than the others which showed AD values according to the data observed in mainland foci [30] , [31] . Factors such as rural condition , a more isolated location , a suitable environmental for resting and breeding ( cocoa and coffee plantations ) and the absence of vector control activities may have contributed to the spreading of tsetse fly populations in these villages . Moreover , the presence of wild and domestic animal hosts in this remote area of the southern Luba district could be the most conditioning factor in the vector densities . A higher sex ratio ( female/male ) of tsetse flies was also noticed in B . Drumen and B . Avendaño while this rate was near 50% in the other villages sampled . In previous studies carried out in mainland foci , a seasonal variation of sex ratio was observed , being higher after a peak of G . p . palpalis density typically reached at the end of rainy seasons [30] , [31] . The sampling of this study was carried out during that period in order to collect as many flies as possible for a more accurate analysis and then , a high female/male ratio was expected . This pattern is more noticeable in the two villages with a higher density of vector where population seems to be well established and follow the natural dynamic . On the other hand , the individuals caught in the second week were younger , suggesting that the removal of flies during the first week had an impact over population dynamic . This phenomenon was also noticed in previous studies in this focus [26] , [27] . It is consistent with the known low reproductive rate of Glossina genus and its relatively small population sizes [34] , [51] , [52] . Vector control was performed in Luba during a few months in order to assess its utility for reducing the human – vector contact . It was later given up and control activities were focused in active screening of human cases and passive detection in the district hospital [5] , [27] . The high cost and maintained efforts needed for vector control strategies make it unsuitable in conditions of very low human infection rate . Nevertheless , strategies such as active screening , chemoprophylaxis with drugs , treatment of infection or spraying animals with insecticides to prevent bites of tsetse flies would not be viable in a wild cycle and hence , vector control would be the only option for an indirect intervention at this level . A vector control campaign focused in the areas surrounding villages with higher density of tsetse fly could reduce the vector - fauna contact , enabling a permanent elimination of the parasite in the epidemiological cycle . In the last two decades , successful control campaigns have been carried out in Luba . Its insular situation gives a degree of isolation which makes more difficult the reintroduction of new cases or infected vectors from neighbouring countries , condition which allowed the tsetse fly elimination in others islands such as Zanzibar and Principe [53]–[55] . Other foci in the mainland Equatorial Guinea , where the same control activities were undertaken , have showed a fall of reported patients but T . b . gambiense infection was never completely cleared and a constant drop of cases per year is currently being described [3] . Other factors , such as economical changes ( mainly petroleum exploitation ) and the subsequent abandon of rural activities such livestock breeding and agriculture , could have contributed to the exceptional success of these control campaigns in Luba . The results of this study suggest interesting features about trypanosomiasis epidemiology in Luba focus . Several differences have been noticed with regards to the other foci of Equatorial Guinea . T . brucei s . l . prevalence in domestic animals is much lower in Luba and no positive T . b . gambiense samples were found . By contrast , T . brucei s . l . infection rate in tsetse flies was high which could be a signal of an intense transmission . Taking into account both data , the hypothesis of the wild fauna as an important feeding source of Glossina spp . and T . brucei s . l reservoir should be considered . Although the prevalence rate is very low , T . b . gambiense infection in tsetse fly also confirms the theory of the permanence of this parasite in Luba focus . It could be concluded that controlling HAT in a given focus is a complex aim and different approaches must be addressed; conventional active human screening is an efficient strategy to decrease the number of cases but other interventions ( such as vectorial control and management of other reservoirs ) could be assessed in order to ensure the elimination of the parasite . In the past , Luba suffered the effects of the neglect of successful control activities leading to a resurgence beginning the 1980s after more than 20 years of apparent absence of the parasite [26] . Nowadays , with the improved epidemiological knowledge achieved by decades of experience fighting the sleeping sickness , resurgences of this disease could be avoidable .
|
Sleeping sickness is a neglected disease with an important impact on public health of many countries of Sub-Saharan Africa . It is transmitted by tsetse fly bites ( the vector ) and mainly affects remote and rural populations . The chronic form , caused by Trypanosoma brucei gambiense , includes almost 90% of reported cases , and it is often misdiagnosed or lately detected after months or years of infection . Many efforts have been carried out to control the disease and interesting advances have been achieved . Although elimination is considered possible , there is an urgent need to understand the disease dynamic , especially in foci with very low rate ( or absent ) of infection for a long time . We performed a parasite screening in tsetse flies and livestock from Luba focus ( Equatorial Guinea ) , considered to be “controlled” since 1995 ( no human cases for fifteen years ) . The obtained results demonstrate that T . b . gambiense still remains in the environment and entomological data reveal high population density of the vector in some localities . This finding suggests that other intervention ways focused on control of vector populations , combined with the detection of human cases , could be necessary to achieve the total elimination of the parasite in hypoendemic foci .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"public",
"health",
"and",
"epidemiology/infectious",
"diseases",
"public",
"health",
"and",
"epidemiology/screening",
"infectious",
"diseases/protozoal",
"infections",
"infectious",
"diseases/epidemiology",
"and",
"control",
"of",
"infectious",
"diseases"
] |
2010
|
Screening of Trypanosoma brucei gambiense in Domestic Livestock and Tsetse Flies from an Insular Endemic Focus (Luba, Equatorial Guinea)
|
Both environmental factors and genetic loci have been associated with coronary artery disease ( CAD ) , however gene-gene and gene-environment interactions that might identify molecular mechanisms of risk are not easily studied by human genetic approaches . We have previously identified the transcription factor TCF21 as the causal CAD gene at 6q23 . 2 and characterized its downstream transcriptional network that is enriched for CAD GWAS genes . Here we investigate the hypothesis that TCF21 interacts with a downstream target gene , the aryl hydrocarbon receptor ( AHR ) , a ligand-activated transcription factor that mediates the cellular response to environmental contaminants , including dioxin and polycyclic aromatic hydrocarbons ( e . g . , tobacco smoke ) . Perturbation of TCF21 expression in human coronary artery smooth muscle cells ( HCASMC ) revealed that TCF21 promotes expression of AHR , its heterodimerization partner ARNT , and cooperates with these factors to upregulate a number of inflammatory downstream disease related genes including IL1A , MMP1 , and CYP1A1 . TCF21 was shown to bind in AHR , ARNT and downstream target gene loci , and co-localization was noted for AHR-ARNT and TCF21 binding sites genome-wide in regions of HCASMC open chromatin . These regions of co-localization were found to be enriched for GWAS signals associated with cardio-metabolic as well as chronic inflammatory disease phenotypes . Finally , we show that similar to TCF21 , AHR gene expression is increased in atherosclerotic lesions in mice in vivo using laser capture microdissection , and AHR protein is localized in human carotid atherosclerotic lesions where it is associated with protein kinases with a critical role in innate immune response . These data suggest that TCF21 can cooperate with AHR to activate an inflammatory gene expression program that is exacerbated by environmental stimuli , and may contribute to the overall risk for CAD .
Genome-wide association studies ( GWAS ) have identified susceptibility loci and candidate genetic variants that predispose to atherosclerotic coronary artery disease ( CAD ) in humans . [1–4] Despite significant advances made in mapping the genetic contribution to CAD , there has been limited progress toward understanding molecular mechanisms leading to increased atherosclerosis susceptibility that are mediated through gene-environment ( GxE ) interactions . [5] The difficulty in identifying the role of genetic variation in the differential response to environmental exposure stems from inaccurate quantification of the exposure , the inability to isolate the exposures of interest , and the lack of statistical power . [6] GWA studies have identified variation at 6q23 . 2 to be associated with CAD in Caucasian and Han Chinese populations[1 , 7] , and work in this lab has identified TCF21 as the causal gene in this locus . [8 , 9] Mechanistic studies employing lineage tracing in murine disease models have found that Tcf21 expression is localized to the medial and adventitial layers of the coronary vessel wall at baseline , and that Tcf21 expressing cells migrate through the lesion and contribute to the fibrous cap as disease progresses . [10] These data , in combination with in vitro studies indicating that TCF21 inhibits differentiation and promotes SMC proliferation , suggest a role for this transcription factor in the phenotypic modulation of medial SMC in the response to vascular injury . [10 , 11] Further , our RNA-seq and ChIP-seq studies have shown that TCF21 binds and regulates a network of genes associated with CAD . [12] We discovered one of the central components of the TCF21 gene network to be the aryl hydrocarbon receptor ( AHR ) , a transcription factor that mediates the response to environmental toxins and xenobiotics , and is known to regulate the inflammatory cellular response . [13–17] AHR binds to a complex array of nuclear proteins involved in diverse processes related to signaling through hormone receptor and inflammatory pathways , chromatin remodeling , etc . , and activates a number of target genes , including cytochromes P450 ( CYP1A1 and CYP1B1 ) , and AHR repressor ( AHRR ) . [18 , 19] AHR is active primarily in the liver , however it is also strongly expressed in the cardiovascular system , where it has been described to play a role in the cardiovascular development and vascular remodeling . [20–22] In the context of environmental stimuli , AHR ligands include a wide range of environmental pollutants , including 2 , 3 , 7 , 8-tetrachlorodibenzo-p-dioxin ( dioxin ) , co-planar polychlorinated biphenyls ( PCBs ) , and polycyclic aromatic hydrocarbons ( PAH ) which are major constituents of tobacco smoke . [15 , 22–24] The correlation of major cardiovascular risk factors with the AHR pathway relates to epidemiological evidence that dioxin exposure is linked to increased cardiovascular mortality . [25] Furthermore , murine model studies have shown that mice carrying an AHR variant with higher ligand affinity developed more severe atherosclerosis compared to wild-type mice[22] , and an increase in disease burden when exposed to dioxin . [26] In humans , a common SNP associated with AHR was found to correlate with the CAD phenotype in a Chinese population . [27] In addition , the expression level of AHR in circulating peripheral mononuclear cells was associated with acute coronary syndromes ( ACS ) , suggesting that greater AHR level might be associated with plaque instability and rupture . Given the role of AHR in mediating inflammation and atherosclerosis , we postulated that TCF21 may alter the risk of atherosclerosis by modulating the AHR pathway . We set out to characterize the intersection of these pathways at the genomic level , identify the possible mechanisms of interaction , and to determine the role of TCF21-AHR interactions in the context of inflammation in the vessel wall . Through these studies , we define the molecular mechanisms by which these two transcriptional pathways interact to regulate the risk of atherosclerosis .
We have previously reported the analysis of an in vitro siTCF21 knockdown RNA-seq study in human coronary artery smooth muscle cells ( HCASMC ) and noted differential expression of a number of inflammatory genes and pathways . [10] Interestingly , this module included the gene encoding AHR which directs an inflammatory program as part of its repertoire of response to xenobiotics[13 , 14 , 26] , and the xenobiotic pathway was identified as one of those differentially regulated with TCF21 modulation . [10] Also , ChIP-seq studies in HCASMC have shown TCF21 binding in the AHR locus , suggesting that this gene is regulated in part by TCF21 , and raising our interest in possible interactions between these transcriptional networks . [12] While AHR has not been associated with CAD risk using the statistical criteria employed for GWAS efforts , we have identified a variant within the AHR locus ( rs608646 ) that has a nominal association ( p = 0 . 0047 ) in the CARDIOGRAM+C4D GWAS data in the context of a single SNP study ( S1A Fig ) . This SNP was also noted to regulate expression of AHR as identified with eQTL studies in GTEx tissues high in SMC content ( S1B Fig ) . In aortic and coronary artery tissues , the AHR locus ( +/- 1Mb ) was generally enriched with AHR-eQTL signals with multiple peaks uniformly scattered throughout the locus ( S2 Fig ) , indicating that AHR gene expression is genetically regulated in the vasculature and emphasizing the importance of AHR in these tissues . These data thus validate a previous candidate gene study that found association of CAD with AHR in East Asians . [26] To further investigate overlap of these pathways , we sought to expand the repertoire of TCF21 regulated genes by performing a transcriptome analysis with lentivirus mediated TCF21 over-expression in HCASMC . Analysis of the top 500 differentially regulated genes with goseq[28] ( Fig 1A , S1 Table ) identified a significant number of terms related to embryonic development ( lung morphogenesis , lung vasculature development ) , but also numerous terms related to innate immunity and inflammation ( response to bacterial lipopeptide , response to lipoteichoic acid , CCR chemokine receptor binding , chemokine receptor binding , lymphocyte chemotaxis , chronic inflammatory response ) , most of which were upregulated with TCF21 overexpression . Employing the DAVID algorithm we found that downregulated genes were primarily associated with SMC development and phenotype ( regulation of blood vessel size , contractile fiber/myofibril ) while upregulated genes were primarily associated with cellular proliferation ( mitosis/cell cycle , positive regulation of DNA replication ) ( Fig 1B , S2 Table ) . In addition , GO terms enrichment and PCA analysis performed using goseq yielded multiple immune system and atherosclerosis- related GO terms , implicating TCF21 in HCASMC to promote immune and pro-atherosclerotic-responses ( S3 Fig ) . To look for relationships between AHR and TCF21 transcriptional networks , we investigated correlations among genes that reside in the co-expression modules of both TCF21 and AHR , using publicly available microarray data sets . We created genome-wide gene expression modules using 4164 human microarray data sets , and used the GeneFriends algorithm that reports the top 5% of co-expressed genes as high order associations , as well as second order indirect , associations . [29] In this analysis , TCF21 and AHR shared a number of indirect associations that link the two modules ( Fig 1C ) . Gene ontology analysis of associations with p<0 . 05 between TCF21 and AHR networks revealed strong enrichment for inflammation , extracellular matrix modification , and developmental terms ( S3 Table ) . TCF21 and AHR appeared to be highly related when co-expression network was visualized with all other CAD GWAS implicated genes , localizing in a cluster of extracellular matrix gene COL4A1 and growth factor receptor PDGFR , and distinct from a cluster of lipid genes ( LPA , APOA5 , APOB , APOA1 ) ( Fig 1D ) . Further , we identified the co-expression module for ARNT , the heterodimer partner of AHR , and found that it also contains genes indirectly connected to AHR and TCF21 modules ( S4 Fig ) , suggesting functional connectivity between AHR-ARNT and TCF21 co-expressed genes ( S3 Table ) . Experiments were first conducted to determine whether TCF21 modulates expression of AHR and ARNT . RNA-seq and qPCR analysis in HCASMC showed that the AHR and ARNT mRNA levels were down-regulated by TCF21 siRNA knockdown ( AHR 1 . 0±0 . 07 vs . 0 . 58±0 . 02 , p = 0 . 0037; and ARNT 1 . 0±0 . 04 vs . 0 . 63±0 . 04 , p = 0 . 0031 ) ( Fig 2A and 2B ) , and up-regulated by TCF21 overexpression ( S5 Fig ) . To further characterize the intersection of TCF21 and AHR transcriptional networks , we investigated the mechanism by which TCF21 regulates downstream genes in the AHR pathway , and chose to first study the dioxin effect on the canonical AHR target gene CYP1A1 . Dioxin induction of CYP1A1 mRNA levels nearly doubled in HCASMC exposed to both dioxin and TCF21 transfection compared to dioxin alone ( 153 . 5±9 . 7 vs . 61 . 8±5 . 9 fold , P = 0 . 001 ) , and the opposite result was seen with TCF21 knock-down in conjunction with dioxin ( 99 . 5±29 . 8 vs . 247 . 9±64 . 8 , P<0 . 05 ) ( Fig 2C and 2D ) . Manipulation of TCF21 expression alone did not alter CYP1A1 expression , suggesting that it does not directly affect transcription of this canonical AHR downstream gene , but does alter the response to dioxin most likely through regulation of AHR and ARNT expression levels . To investigate the possible interaction of TCF21 and AHR in the regulation of target inflammatory pathway genes , additional studies were conducted in HCASMC . We focused on IL1A and MMP1 genes , as previous studies have found these genes to be representative targets of AHR activation through direct or indirect pathways . [30–32] mRNA levels were measured for IL1A , and MMP1 genes by RT-PCR in HCASMC with TCF21 expression perturbed by knockdown and over-expression . Knockdown of TCF21 decreased IL1A expression compared to cells treated with scrambled siRNA ( 0 . 91±0 . 04 vs . 0 . 52±0 . 06 , p = 0 . 013 ) ( Fig 2E ) . Dioxin treatment significantly increased expression of IL1A ( 0 . 91±0 . 04 vs . 1 . 42±0 . 09 , p = 0 . 007 ) , and co-treatment of siTCF21 blocked this effect ( 1 . 42±0 . 09 vs 0 . 65±0 . 04 , p = 0 . 001 ) . Similar results were obtained for MMP1 , with siTCF21 alone decreasing gene expression ( 1 . 02±0 . 06 vs . 0 . 65±0 . 05 , p = 0 . 009 ) , dioxin increasing expression ( 1 . 02±0 . 06 vs . 1 . 62±0 . 28 , p = 0 . 10 ) , and siTCF21 knocking down the increased expression of MMP1 seen with dioxin ( 1 . 62±0 . 28 vs . 0 . 64±0 . 12 , p = 0 . 034 ) ( Fig 2F ) . To begin to investigate the mechanism by which TCF21 regulates AHR expression , we correlated whole genome RNA-seq and genotype information developed in 52 HCASMC lines to evaluate expression quantitative trait locus ( eQTL ) effects at the AHR locus . [33] We identified SNP rs10265174 to be one of the top eQTLs for AHR ( p<9e-5 ) ( Fig 3A and 3B ) . Also , rs10265174 was consistently found to regulate gene expression in multiple GTEx tissues , including coronary artery , aorta and tibial artery ( S4 Table ) . Furthermore , the SNP was located in an open chromatin region/enhancer marked by ATAC-seq , H3K27ac ChIP-Seq , JUN and JUND ChIP-Seq peaks , and within a TCF21 ChIP-Seq peak . We found the rs10265174 variant to alter the PMW scores for AP1 and TCF4 transcription factors ( HaploReg ) ( Fig 3C ) . Given that TCF4 is a known bHLH binding partner for TCF21 , [34] we evaluated whether TCF21 might directly regulate AHR gene expression at this site . We surveyed ChIP-seq data previously generated for TCF21 in HCASMC[12] and identified ChIP-seq peaks representing TCF21 binding sites in both the AHR and ARNT genes . We confirmed the binding of TCF21 to both genomic regions in HCASMC with ChIP-qPCR for AHR ( 1 . 02±0 . 29 vs . 3 . 58±0 . 75; p = 0 . 033 ) and ARNT ( 1 . 03±0 . 28 vs . 20 . 04±4 . 80; p = 0 . 017 ) loci ( Fig 3D , S6 Fig ) . Taken together , these data suggest that TCF21 may directly regulate expression of both the AHR and ARNT genes at the transcriptional level . To further investigate the overlap of TCF21 and AHR transcriptional networks at the genomic level , we determined the genome-wide relationship between binding sites for AHR-ARNT and TCF21 . We scanned the human genome sequence with the position weight matrix ( PWM ) for AHR-ARNT , and scanned for the PWM for TCF12 as a surrogate for TCF21 ( JASPAR matrices , Ahr::Arnt—MA0006 . 1; Tcf12—MA0521 . 1 ) . [12 , 35 , 36] TCF12 is the primary heterodimerization partner for TCF21 , binds the same primary sequence as TCF21 , and this composite site is indicated here as TCF12/TCF21 . [12 , 36] Predicted binding sites for TCF12/TCF21 and AHR-ARNT identified co-localization within a broad region of 5kb , with 339 high stringency TCF12/TCF21 and AHR-ARNT sites that directly overlap ( P<2 . 2e-16 , Fisher exact test , using combined ENCODE open chromatin regions as background; 218 sites overlapping within the background ) ( Fig 4A and 4B ) and 11769 lower stringency sites overlapping ( P<2 . 2e-16 , Fisher exact test , ENCODE background; 4833 sites overlapping within the background; S5 Table ) . Next , we tested the positional orientation of TCF12/TCF21 and AHR-ARNT sites near functional elements , such as promoters , using the collection of 100 , 276 human ENSEMBL transcription start sites ( TSS ) , Hum_ENSEMBL69 from Biomart . We observed that both matrices show double peaks near the oriented ENSEMBL TSS ( Fig 4C ) . We also noted that these double peaks are in phase with each other , suggesting conservation of spatial orientation between the two sets of predicted binding sites and possible functional interaction of the two proteins that is preserved by evolutionary constraint near the TSS . In addition , we observe that the distance between phased peaks corresponds to the position of the +1 nucleosome , with an additional peak corresponding to the +2 nucleosome in the TCF PWM profile , suggesting that functional interaction between TCF21 and AHR-ARNT would be localized at the boundaries defined by nucleosome positioning at functional regions . To test whether TCF21 in vivo binding sites correlate with AHR-ARNT PWM predictions , we generated the precise locations of TCF21 ChIP-seq summit positions in HCASMC using the MACS ChIP-seq tool ( S6 Table ) . [37] We found that the center of these TCF21 ChIP-seq summits co-localized with the predicted AHR-ARNT PWM sites , further suggesting that AHR-ARNT complexes co-localize genome-wide with TCF21 in vivo binding sites ( Fig 4D ) . In contrast , control PWMs for kidney/liver specific factors HNF1A and HNF1B showed uniform background distribution near TCF21 summits ( Fig 4E and S7 Fig ) . In addition , AHR-ARNT PWM profiles showed an increase at summits for open chromatin regions in HCASMC , defined with MACS and ATAC-seq HCASMC data sets ( Fig 4F ) . Control HNF1A and HNF1B matrices showed a decrease in their frequency near ATAC-seq summits ( Fig 4G and S7 Fig ) . We further assessed the steric relationship of TCF21 and AHR binding by dividing the co-localized TCF and AHR predicted sites into two categories , rotationally phased and un-phased , as rotational phasing has been shown to be crucial for direct protein binding . [38 , 39] We considered PWM sites to be phased if they occurred at distances of n ( 10bp ) , i . e . 10 , 20 , 30 , and 40bp , in which case due to the DNA helical pitch they will be oriented on the same side of the DNA strand and capable of direct protein-protein interaction . If separated by distances of 5 , 15 , 25 , 35 , 45 bp they would be expected to be oriented on the opposite sides of the DNA molecule due to the pitch of the DNA major groove , rotationally un-phased and less capable of direct protein-protein interaction . In the case of un-phased sites , indirect interaction through a protein complex might still be possible , e . g . through intermediate protein interactions . We extracted genes that are proximal to both categories of sites and calculated GO enrichment using GREAT ( Fig 5 , S7 Table ) . [40] The un-phased sites were enriched in cellular differentiation categories such as: negative regulation of cell fate commitment , but importantly a number of terms were related to inflammatory response ( regulation of cytokine production , regulation of interleukin-6 production , regulation of TNF production ) . GO terms for phased sites were enriched in cellular and developmental terms ( skeletal system morphogenesis , response to organophosphorous , regulation of cell migration , cell-matrix adhesion , and osteoblast development ) In addition , binomial fold changes for un-phased AHR-TCF sites were ~10 times higher than for phased sites , implicating the indirect interaction of AHR-TCF21 factors as predominant in the human genome . To further evaluate whether TCF21 and AHR-ARNT complex binding co-localizes in the human genome , we compared ChIP-seq data for these three TFs . We identified TCF21 ChIP-seq peaks in HCASMC that overlapped with AHR and ARNT ChIP-seq sites identified in MCF-7 cells[41] , and obtained a statistically significant co-localization of sites using the Fisher’s exact test . Overlap of TCF21 and AHR peaks produced odds ratios within confidence intervals CI: 4 . 34–5 . 4 ( p = 1 . 88e-121 ) , for ARNT and TCF21 , CI: 4 . 16–5 . 77 ( p = 1 . 7e-56 ) and for AHR , ANRT and TCF21 , CI: 4 . 91–7 . 05 ( p = 1 . 13e-56 ) ( S8 Table ) . We obtained in total 322 ( 12 . 4% ) genomic locations that were co-occupied by AHR and TCF21 , out of which 119 sites were also occupied by ARNT . Similarly , in 143 genomic locations ( 10 . 5% ) ARNT co-localized with TCF21 , out of which 119 were occupied with its binding partner AHR ( Fig 6A ) . Overlap of AHR and ARNT identified 890 sites ( AHR , 34 . 3%; ARNT , 65 . 8% percent of total sites ) , consistent with the fact the two proteins are known binding partners . Subsequently , we selected genes that are proximal to the overlapping TCF21-AHR , TCF21-ARNT and TCF21-AHR-ARNT ChIP-seq sites , and performed GO enrichment analysis with GREAT ( Fig 6B , S8 Table ) . TCF21 and AHR-ARNT overlapping sites classified into GO-terms related to chemokine and cytokine signaling ( positive regulation of cytosolic calcium ion concentration , regulation of cytosolic calcium ion concentration ) , apoptosis ( regulation of apoptotic process , regulation of programmed cell death ) , metabolic processes ( cellular hormone metabolic process , isoprenoid metabolic process ) , and cellular signaling ( cellular response to stimulus ) . Next , we assessed the binding of AHR , ARNT and TCF21 near lead SNPs from the GWAS Catalog ( version 2016-05-08 ) , expanded by addition of CARDIoGRAM+C4D meta-analysis data[2 , 42] , using the binomial test for genomic overlap . We scanned the lead SNPs using windows of +/-2kb , +/-5kb , and +/-10kb near the ChIP-seq binding sites and calculated the significance using the gwasanalytics tool ( Fig 6C–6F , S8 Fig ) . Using the +/- 2kb window , TCF21 binding shows general enrichment near a wide range of GWAS SNPs for cardio-metabolic phenotypes ( coronary heart disease , blood pressure and type 2 diabetes ) as well as chronic inflammatory diseases ( Crohns disease , multiple sclerosis , and rheumatoid arthritis ) , as well as skeletal phenotypes ( bone mineral density ) and in certain neurological disorders ( bipolar disorder and schizophrenia ) . ARNT binding was localized near GWAS SNPs for chronic inflammatory diseases ( lupus erythematosus and ulcerative colitis ) and prostate cancer GWAS SNPs . AHR binding co-localized with CAD variants ( coronary artery disease , coronary artery calcification ) as well as chronic inflammatory GWAS SNPs ( e . g . , Crohn’s disease ) . After intersection of AHR with ARNT and TCF21 the only remaining categories were coronary artery disease and coronary artery calcification , narrowing the importance of the interaction of AHR/ARNT and TCF21 factors to pathophysiological processes in cardiovascular disease . Furthermore , we surveyed the overlap of CARDIoGRAM+C4D GWAS SNPs and AHR-ARNT PWM to consider the potential role of AHR in other CAD associated genes . In total 456 ARNT-AHR sites overlapped with CARDIOGRAM+C4D SNPs ( lead plus LD r2>0 . 8 ) , comprising 0 . 27 permil ( low stringency ) and 0 . 38 permil ( high stringency ) of total ARNT-AHR PWMs . In comparison , there were only 7 and 5 HNF1A and HNF1B sites , comprising 0 . 12/0 . 10 permil of total HNF1A/B sites ( p<0 . 005 , comparison of AHR-ARNT and HNF using Z-score test for proportions , S9 Fig ) . Given these data showing that TCF21 and AHR binding sites are co-localized in the genome ( Figs 4–6 ) , and that TCF21 expression levels directly modulate the AHR response to dioxin ( Fig 2 ) , we investigated the functional interaction of these transcription factors at target loci . First , we surveyed the genomic region of CYP1A1 for TCF21 in vivo binding in HCASMC . A TCF21 ChIP-seq binding peak was identified and localized to a region of open chromatin , as defined by ATAC-seq data in HCASMC[10] , and binding was confirmed with ChIP-qPCR ( IgG 1 . 0±0 . 18 vs . TCF21 4 . 71±0 . 24 , p = 0 . 001 ) ( Fig 7A and 7B ) . This peak co-localized in the same region of open chromatin with several predicted AHR-ARNT binding sites , thus suggesting coordinated regulation of CYP1A1 expression . To confirm that the regulation of CYP1A1 mRNA levels by AHR and TCF21 is mediated at the transcriptional level through the observed ChIP identified binding sites , and to look for evidence of cooperativity at this level , we conducted reporter gene transfection studies . A 50 bp sequence containing alternating binding motifs for TCF21 and AHR-ARNT binding identified in the CYP1A1 gene was cloned into a luciferase expression plasmid with a minimal promoter sequence . Dual luciferase-renilla assays revealed enhancer activity when exposed to TCF21 overexpression ( 1 . 0±0 . 06 vs . 3 . 15±0 . 02 , p = 0 . 0024 ) , and TCF21 overexpression further increased the luciferase expression induced by dioxin in these cells ( 3 . 62±0 . 30 vs . 6 . 17±1 . 31 fold , p = 0 . 0005 ) ( Fig 7C ) . The combined effect was additive with no evidence of synergism that would be suggestive of cooperative binding . Further , when the TCF21 binding motifs were removed from the reporter construct , TCF21 overexpression failed to further increase the expression of luciferase , suggesting that the transcriptional effect of TCF21 is specific for protein-DNA binding ( Fig 7D ) . These data indicate that the regulatory effect of TCF21 on AHR target genes can be mediated by direct interaction in these target loci , and requires protein-DNA binding . We followed up our previous studies showing regulation of inflammatory mediators by AHR and TCF21 with studies investigating possible endogenous mediators of AHR activation . As shown previously , application of dioxin to HCASMC resulted in up-regulation of IL1A , and this effect was reversed when cells were treated with the AHR antagonist alpha-napthoflavone ( α-NF ) ( Fig 8A ) . In the same experiments , we tested oxidized-LDL ( ox-LDL ) as a potential endogenous activator of AHR in HCASMC . [43 , 44] The treatment with ox-LDL resulted in the activation of genes that was similarly reduced with α-NF co-treatment , suggesting that the SMC response to ox-LDL is at least partly mediated by the AHR pathway ( Fig 8A ) . We also found activation of a dioxin response element with oxLDL in luciferase assays ( S10 Fig ) . Given these data suggesting that AHR targets overlap the TCF21 CAD associated transcriptional network genes , we sought to substantiate the relevance of AHR in vascular disease through expression studies in mouse and human vascular tissues[8–10] For in vivo gene expression in mice , we performed microarray analysis of carotid arteries subjected to plaque rupture induced by partial ligation in ApoE-/- animals to compare gene expression between ruptured and non-ruptured plaques . [45] We found Ahr expression to be higher in the ruptured plaques compared to non-ruptured plaques ( 8 . 60±0 . 20 vs . 9 . 58±0 . 16 , FDR q = 0 . 054 ) ( Fig 8B ) . Furthermore , laser capture microdissection ( LCM ) was performed in atherosclerotic lesions in the aortic sinus of ApoE-/- mice exposed to 12 weeks of high fat diet . We found the expression level of Ahr to be significantly higher in the intimal plaque when compared to the expression in the adventitia , localizing the expression of AHR to the pathologic intimal thickening ( 1 . 0±0 . 2 vs . 12 . 4±1 . 2 , p = 0 . 0008 ) ( Fig 8C ) . Next , we validated these findings in human arteries ex vivo , using microarray based expression data from normal arteries and atherosclerotic human carotid lesions from the BiKE repository . [46] Expression levels of AHR along with IL1A , and MMP1 were significantly higher in the diseased lesions ( Fig 8D , AHR 8 . 71±0 . 16 in normal vs . 9 . 27±0 . 05 in plaques , p = 0 . 0065; S11 Fig ) . Furthermore , we analyzed the proteins present in human carotid plaques using liquid chromatography tandem mass-spectrometry ( LC-MS/MS ) . Proteomic datasets were constructed from highly phenotyped patients with asymptomatic and symptomatic carotid stenoses , 10 subjects each matched for gender , statin usage and age , with plaques selected on CT and histology criteria . AHR-TCF21 unique interactors from BIOGRID protein-protein interaction database were used to display clustering patterns ( Fig 8E ) . AHR is located in a cluster of genes that include immune related genes such as IRAK4 , interleukin-1 receptor-associated kinase 4 , transcription factors like SP1 , and cell cycle regulated genes including XPO1 . In addition , we selected ChIP-seq co-occupied genes for AHR-TCF21 and AHR-ARNT-TCF21 transcription factors and observed clustering of AHR target protein CYP1B1 with extracellular matrix factors FN1 , COL18A1 and with growth factor receptor IGF1R , implicating AHR and its downstream targets in regulation of extracellular matrix component of the diseased human carotid artery plaque ( S12 Fig ) .
We have identified TCF21 as the causal gene at 6q23 . 2 , characterized its mechanism of association , and shown that binding of this transcription factor is enriched in other CAD associated loci . [8 , 12] To investigate how TCF21 interaction with other CAD loci may regulate disease risk , we have begun to study mechanisms of association in these loci . For initial studies we have chosen the AHR gene , because it encodes a transcription factor , allowing direct study of its downstream signaling pathway , and because of the well-characterized link between this factor and environmental exposures that are relevant for cardiovascular disease . This work thus addresses two aspects of CAD that have not been directly approachable with human association studies , investigating both gene-by-gene and gene-by-environment contributions to disease genetic risk . Although variants in the AHR locus ( rs608646 ) have shown only nominal association with CAD risk in GWAS meta-analyses , this may be due to the technical limitation of the GWAS methodology in the AHR locus or inadequate statistical power . It remains possible if not likely that AHR functions as a hub or master regulator in CAD without harboring regulatory disease variants . We did identify a variant within the AHR locus ( rs608646 ) that has a moderate association ( p = 0 . 0047 ) in the CARDIOGRAM+C4D GWAS data ( S1A Fig ) , and this SNP was also noted to regulate expression of AHR as identified with eQTL studies in GTEx tissues high in SMC content ( S1B Fig ) . In aortic and coronary artery tissues , the AHR locus was enriched with AHR-eQTL signals with multiple peaks across the genomic region ( Fig 3B , and S2 Fig ) , indicating that AHR gene expression is genetically regulated in the vasculature and emphasizing the relevance of AHR expression in these tissues . Further , we also found genome-wide enrichment of the PWM for AHR-ARNT within CARDIOGRAM+C4D GWAS loci ( S9 Fig ) , suggesting that the effect of AHR on CAD may be partly via genetic variation in protein-DNA interaction near genes related to CAD . These data support the candidate gene study which found an association of CAD with the AHR locus in East Asians . [26] In our studies , we have pursued numerous approaches to investigate links between these two genes and their related transcriptional networks , and to investigate mechanisms by which they may work together to modulate CAD risk . First , we have shown with targeted studies that TCF21 binds both the AHR and ARNT loci , and increases expression levels of these genes in HCASMC , confirming previously published genomic studies and RNA-seq studies reported here . Second , these studies provide evidence for overlap of the TCF21 and AHR transcriptional networks . Both TCF21 and dioxin were shown to increase expression of disease-related factors such as IL1A , MMP1 and interestingly knockdown of TCF21 was able to almost completely abolish the effect of dioxin , suggesting that the inflammatory activation by AHR is dependent on the presence of TCF21 . AHR is well known to promote inflammation in a number of situations , and to work with NFkB in this regard . [17] Also , we have previously shown that TCF21 can promote expression of a number of inflammatory genes and we show that this pro-inflammatory program represents an intersection of TCF21 with AHR function , identifying a subset of TCF21 target genes that could create a highly inflammatory cellular profile that would be significantly magnified with relevant environmental exposures . We also found that oxidized LDL activated the AHR pathway in HCASMC , consistent with previous reports in other cell types . [43 , 47] Further analyses investigated additional mechanisms of interaction between these two pathways . Using PWMs for both TCF21 and AHR , we found highly significant enrichment for co-localization in regions of open chromatin in HCASMC , and characterized similar organization of these binding sites around transcription factor start sites , suggesting functional interaction between TCF21 and AHR and the basal transcriptional apparatus , as proposed previously for other TFs . [48 , 49] The genomic co-localization was further refined by intersecting summit locations from TCF21 ChIP-seq data with AHR-ARNT PWM positions . Support for these observations reflecting in vivo associations was provided by co-localization of ChIP-seq peaks for TCF21 , AHR , and ARNT . These data suggest a role for AHR-ARNT in the functional regulation of coronary SMC phenotype . Co-localization of TF binding often suggests direct functional interaction , and since TCF21 and AHR may regulate transcription in the same direction , an obvious hypothesis is that they bind cooperatively either through direct protein-protein interaction or through joint recruitment of ancillary adaptor proteins . [50] In addition to the genomic co-localization data , the absence of IL1A response to dioxin in TCF21 knockdown , and our studies investigating the phasing of binding site placement also suggests some form of direct or indirect molecular interaction . The striking difference between functional annotations for the two categories of steric relationship are consistent with different functional interactions between AHR–ARNT and TCF21 in the context of DNA binding . GO terms for un-phased sites showed much stronger enrichment and significance compared to those of the phased sites , supporting indirect interaction as the likely functional mechanism . We investigated this possibility using the CYP1A1 gene as model locus where both transcription factors bind . The reporter gene transfection studies with constructs containing both TCF21 and AHR binding sites showed an additive effect suggesting that the transcriptional effects are due to each TF acting independently . However , both TCF21 and AHR are known to bind AR[51 , 52] , and also potentially bind to Rb1 ( S13 Fig ) . It remains a possibility that they interact through other TFs that function as intermediaries and are either not expressed or not active in the HCASMC . In fact , we were not able to demonstrate direct protein-protein interaction using a co-immunoprecipitation assay . Future studies using chemical crosslinking with ChIP and serial ChIP may help resolve these important questions . The functional role of TCF21 in vascular disease appears closely related to its role in embryonic coronary artery vascular development where it is expressed in SMC precursor cells , supporting proliferation and migration of these cells . AHR is also expressed in the developing coronary circulation , and the application of the AHR-ligand dioxin in zebrafish inhibited epicardial and proepicardial development . [53 , 54] Both TCF21 and AHR are downstream of retinoic acid signaling pathways that are critical in coronary artery development . [55–58] The possible molecular interaction of TCF21 and AHR in this setting has not been established . In summary , we describe a novel functional interaction between two bHLH class transcription factors and postulate association of their interaction to the development of atherosclerosis and coronary artery disease ( Fig 9 ) . The discovery of a connection between TCF21 , one of the most highly replicated GWAS candidate genes for coronary artery disease , and AHR , a gene classically involved with response to environmental toxins raises an interesting hypothesis that this interaction may reflect gene-environment interactions that are contributing to CAD and presents an opportunity to define causal gene-gene and gene-environment interactions relevant to the atherosclerotic lesion . Furthermore , our findings that TCF21 and AHR are expressed in the atherosclerotic plaque and that they interact to modulate inflammatory genes and matrix modifying genes suggest that the interaction may directly promote plaque instability leading to myocardial infarction . This work serves to promote mechanistic studies as an approach to understanding gene-by-gene and gene-by-environment contributions to disease genetic risk . Also , once the exact nature of the interaction of the two proteins is fully elucidated , therapeutic targeting of the pro-inflammatory interaction of TCF21 and AHR might be possible to reduce the effects of the pro-atherogenic stimuli in the vasculature and consequently reduce disease risk .
Primary human coronary artery smooth muscle cells ( HCASMC ) were purchased from three different manufacturers , Lonza , PromoCell and Cell Applications and were cultured in complete smooth muscle basal media ( Lonza , #CC-3182 ) according to the manufacturer's instructions . All experiments were performed with HCASMC between passages 5–8 . HEK293 cells were maintained in DMEM containing high glucose , sodium pyruvate and L-glutamine supplemented with 10% FBS . For the TCF21 overexpression study , HCASMC were transduced with 2nd generation lentivirus with TCF21 cDNA cloned into pWPI ( Addgene #12254 ) . Briefly , for lentiviral transduction , the cells were treated at 60% confluence with MOI of 5 for 24 hours . The virus was removed and replaced with low-serum media for 48 hours prior to collection for downstream applications . For the siRNA transfection , cells were grown to 60% confluence , then treated with 10nM siRNA or scramble control with RNAiMax ( Invitrogen , Carlsbad , CA ) for 12 hours . The cells were collected 48 hours after transduction and processed using RNeasy kit ( Qiagen , Hilden , Germany ) for RNA isolation . TCF21 knockdown was performed with siRNA oligos ( Origene , Rockville , MD ) using Lipofectamine RNAiMAX ( Thermo Fisher Scientific , Waltham , MA ) following manufacturer’s protocol . RNA was isolated using RNeasy mini kit ( Qiagen ) and total cDNA was prepared using iScript cDNA synthesis kit ( Biorad , Hercules , CA ) . Gene expression levels were measured using SYBR Green assays with custom designed probes ( S9 Table ) and quantified on a ViiA7 Real-Time PCR system ( Applied Biosystems , Foster City , CA ) and normalized to GAPDH levels . Two group comparisons were performed using student t-test , and three group comparison was performed with ANOVA . HCASMC were cultured as described above and total RNA was purified from 5 . 0x105 cells using the Qiagen miRNeasy kit . RNA libraries were prepared using the Illumina TruSeq library kit as described by the manufacturer . RNA molecules were sequenced using Illumina HiSeq 2500 . Reads contained in raw fastq files were mapped to hg19 using the RNA-seq aligner STAR ( v2 . 4 . 0i ) , that processes data with short run times and yields high numbers of uniquely mapped reads ( https://github . com/alexdobin/STAR ) . Second pass mapping with STAR was then performed using a new index that is created with splice junction information contained in the file SJ . out . tab from the first pass STAR mapping . Read that have been mapped with STAR second pass mapping algorithm were subsequently counted using the htseq-count script distributed with the HTSeq Python package ( https://pypi . python . org/pypi/HTSeq ) . Differential expression of exons , genes , and transcripts were assayed using the DESeq2 R package from Bioconductor ( http://bioconductor . org/packages/release/bioc/html/DESeq2 . html ) , which uses negative binomial distribution to estimate dispersion and model differential expression such as to permit biological variability to be different among tested genes ( transcripts ) . GO terms enrichment and PCA analysis was performed using GOSeq and Gene Set Enrichment and Projection Displays–GSEPD Bioconductor package . Fifty-two human coronary artery smooth muscle cell lines are genotyped using 30X whole-genome sequencing . Genotype calling follows the GATK best practices recommendations . Briefly , after removing adapter with cutadapt , trimmed FASTQ files were aligned with BWA mem , duplicates were marked with Picard tools . After indel realignment and variant base quality recalibration , single-nucleotide variants and short insertion and deletion variants are jointly called on all samples using the GATK Haplotype caller . Called variants are recalibrated and filtered using GATK's variant quality score recalibration module . We used BEAGLE 4 . 1 to impute and phase recalibrated variants using 1000 Genome phase 3 version 5a as a reference panel . After imputation and phasing , we filtered variants based on MAF > 0 . 05 , Hardy-Weinberg equilibrium p-value > 1e-6 , indel length < 51 bps , dosage r2 > 0 . 8 . Gene expression was quantified using mRNA sequencing to an average depth of 50M 75-bp paired-ended reads . Sequences were aligned using STAR two-pass mapping . To avoid allele-specific mapping bias , we removed potentially mismapped reads using WASP . Read counts and FPKM values were generated using RNAseQC . Expression eQTL were mapped with RASQUAL . To remove potential confounders , we included gender , first 3 principal components inferred on the genotypes and first 8 PEER factors inferred on 10 , 000 highest expressed genes . Transcription factor binding and epigenetic annotations of variants were assayed by Haploreg v4 . 1 . A detailed protocol was included in a previous publication [12] . HCASMC were cultured as described above . Antibodies used for ChIP-qPCR were all pre-validated according to ChIP-seq guidelines and ENCODE best practices . Purified rabbit polyclonal antibody against human TCF21 ( HPA013189 ) was purchased from Sigma . Briefly , ChIP-qPCR confirmation was performed using primers designed for the genomic region of AHR , ARNT and CYP1A1 , and compared against ChIP performed with IgG antibody . ( S2 Table ) . qPCR values for AHR , ANRT , and CYP1A1 promoter , normalized relative to the Myogenin ( MYOG ) signal , used as a endogenous control , were expressed as fold change compared to IgG ChIP sample . Comparisons were performed using student t-test . TCF21 ChIP-Seq raw data from Sazonova et al . , were reanalyzed using Model-based Analysis for ChIP-Seq ( MACS v1 . 4 . 2 ) pipeline . Parameters were set to default . Summit locations of the peaks were defined for genome wide correlations with PWM using ChIPCor module—part of ChIP-Seq Analysis Server of the Swiss Institute of Bioinformatics ( ccg . vital-it . ch/chipseq/chip_cor . php ) . TCF21 ChIP-Seq sites were converted to bigwig files and visualized on UCSC Genome Browser . TCF21 overexpression was achieved using cDNA expression construct driven by a CMV promoter transduced by lentivirus . TCF21 knockdown was performed with siRNA oligos ( Origene ) following manufacturer’s protocol . Briefly , for lentiviral transduction , the cells were treated at 60% confluence with MOI of 5 for 24 hours . The virus was removed and replaced with low-serum media for 48 hours prior to collection for downstream applications . For the siRNA transfection , cells were grown to 60% confluence , then treated with 10nM siRNA or scramble control with RNAiMax ( Invitrogen ) for 12 hours . The transduced cells were then treated with TCDD ( Sigma Aldrich Cat#48599 ) at a concentration of 10nM for 24 hours . For the dual luciferase assay , double stranded DNA sequences containing the TCF21 and AHR binding motifs were subcloned into the multiple cloning site ( MCS ) of the pLuc-MCS vector ( Promega , #E1330 ) , located upstream of the translation stop codon and firefly luciferase reporter gene luc2 , driven by the PGK minimal promoter and also carrying the renilla luciferase reporter gene hRluc , as an internal control . Culture media was changed after 6 hrs , and dual luciferase activity was measured after 24 hrs using either SpectraMax L luminometer ( Molecular Devices , Sunnyvale , CA ) . Relative luciferase activity ( firefly/Renilla luciferase ratio ) is represented as the fold change of respective control condition as indicated . Oxidized-LDL was purchased from Alfa Aesar ( Haverhill , MA; Cat No . J65591 ) . Cells were treated at concentration of 10uM for 6 hours with and without α-NF at 10nM ( Sigma Aldrich , St . Louis , MO; Cat No . N5757 ) . The changes in downstream genes were confirmed using RT-qPCR . 12 week old ApoE-/- mice on C56BL/6J background were subjected to 4 weeks of partial ligation followed by 4 days of cuff placement as described previously . [45 , 59] For the aortic sinus atherosclerosis model , ApoE -/- mice were put on 12 weeks of Western high fat diet ( HFD , 21% anhydrous milk fat , 19% casein and 0 . 15% cholesterol , Dyets no . 101511 ) at 4 weeks of age . Using a Leica LMD6000 , we performed LCM of atherosclerotic plaques of mouse aortic sinuses . Briefly , following sacrifice , the cardiac chamber was perfused with PBS , then the aortic sinus was dissected and embedded in Optimal Cutting Temperature ( OCT ) medium ( Tissue-Tek ) . 7um cryosections were placed on to Leica membrane slides , then visualized under the microscope for LCM . Total RNA was extracted using RNeasy Plus Micro kit ( Qiagen ) , and the quality of RNA checked with Agilient Bioanalyzer RNA 6000 Pico kit . The levels of gene expressions were compared using total RNA generated from ApoE ( -/- ) mouse on high fat diet for 12 weeks . Using the published SMART-Seq2 protocol [60] , we amplified the ultra-low input RNA from the LCM . Briefly , reverse transcription was performed using a template switching oligonucleotide ( TSO ) with locked nucleic acid ( LNA ) and Superscript II ( Invitrogen , Carlsbad , CA ) , followed by PCR amplification with KAPA PCR polymerase . Human atherosclerotic carotid artery lesions were obtained from patients undergoing endarterectomy surgery for carotid stenosis , as part of the Biobank of Karolinska Endarterectomies ( BiKE ) . [46] Details of the cohort demographics , sampling at surgery , processing and microarray analyses have been described before . Briefly , normal control samples ( n = 10 ) were iliac arteries and one aorta from healthy organ donors without any history of cardiovascular disease . Plaques were frozen at -80°C immediately after surgery , pulverized to a powder before resuspending in Qiazol lysis reagent ( Qiagen ) and homogenization with a tissue homogenizer . Total RNA was extracted as described above using the miRNeasy Mini Kit ( Qiagen ) and RNA quality assessed using a Bioanalyzer 2100 ( Agilent ) . Global gene expression profiles were analyzed by Affymetrix HG-U133 plus 2 . 0 Genechip microarrays from 127 patient derived plaque samples and 10 donor control samples . Robust multi-array average ( RMA ) normalization was performed and processed gene expression data presented in Log2 scale . Atherosclerotic plaques from 18 BiKE patients ( matched for male gender , age and statin medication ) were analysed using LC-MS/MS as previously described . [46 , 61] Briefly , protein samples were digested by trypsin and the resulting tryptic peptides were TMT-labeled and pooled . Pooled samples were cleaned by Strong Cation exchange columns ( Phenomenex ) and subjected to LC-MS/MS analysis . The sample pools were separated on a 4 hour gradient using an UPLC-system ( Dionex UltiMate™ 3000 ) coupled to a Q-Exactive mass spectrometer ( Thermo Fischer Scientific , San Jose , CA , USA ) . The fragment spectra from the mass spectrometer were matched to a database consisting of theoretical fragment spectra from all human proteins and filtered at a 1% False Discovery Rate ( FDR ) on the peptide level to obtain protein identities ( Uniprot ) . Quantitative information was acquired by using the TMT reporter ion intensities . Correlation matrices were constructed by calculation of the proteomic expression correlation coefficients using the Pearson method and p-values were corrected for multiple comparisons using Bonferroni . For the clustering plots , dissimilarity index was created using the method that best discriminates all correlated pairs , given the formula: Dissimilarity = 1 –Abs ( Correlation ) . Distance matrix was then created from the dissimilarity index . Clustering was performed with heatmap . 2 in gplots . AHR , ARNT , TCF21 , HNF1A , and HNF1B co-expression modules were obtained using GeneFriends using 4164 human microarray data sets or 4133 human RNA-Seq data sets . TCF21 and AHR co-expression modules were defined using 4164 human microarray datasets through GeneFriends and visualized with Cytoscape . TCF21 , AHR and ARNT co-expression modules were defined using 4133 human RNA-seq datasets through GeneFriends and visualized with Cytoscape . GeneFriends associations are defined with the threshold of 5% , meaning the gene is associated to a specific gene if it is in the top 5% co-expressed genes for a that gene . Cytoscape network file was imported for visualization from GeneFriends , containing all gene-gene associations marked as "good friends" ( top 10 friends with a connection strength of 1 ) , or "lesser friends" ( genes ranking between 10 and 20 with a rank of 0 . 5 ) . If a gene is an indirect connection , i . e . friend of a friend , score of 0 . 25 is deduced from the connection strength . Core network of direct interactions is marked on a graph with different colors to distinguish direct from indirect interactions . Connecting co-expression modules and visualization was performed using Cytoscape . Clustering was performed using Edge-weighted Spring Embedded layout . Coronary artery disease ( CAD ) GWAS genes were defined using Cardiogram plus C4D meta-analyses GWAS loci . In total , 77 CAD GWAS genes were used for transcription module analysis in GeneFriends:Microarray to obtain gene expression modules which were subsequently clustered in Cytoskape . Genes TCF21 , AHR , COL4A1 , SMAD3 and PDFGD from the main cluster were colored and indicated , node size was increased and edges to their first neighbors were colored in red . Underlying edge connections were colored purple with increased transparency . Grouping of transcription modules into three main clusters shows that CAD GWAS genes act through three main regulatory networks with TCF21 and AHR gene modules appearing in the single cluster . TCF12 and AHR-ARNT matrices ( TCF12—MA0521 . 1 and AHR-ARNT—MA0006 . 1 ) were obtained from the JASPAR database ( http://jaspar . genereg . net ) [35] . TCF12 PWM was used as it has the same binding motif as TCF21 . Human genome hg19 was scanned with the two JASPAR matrices using PWMScan—Genome-wide PWM scanner ( http://ccg . vital-it . ch/pwmtools/pwmscan . php ) . Position weight matrix sites were counted in windows of various lengths surrounding centered features using the Feature correlation tool from the ChIPCor module from ChIP-Seq Analysis Server ( ccg . vital-it . ch/chipseq/chip_cor . php ) . AHR , ARNT and TCF21 ChIP-Seq binding site were extended to windows +/-1000bp , +/-2000bp , and +/-5000bp using bedtools package . Overlap of extended locations of AHR , ARNT and TCF21 ChIP-Seq binding sites and GWAS Catalog SNPs was performed with bed2GwasCatalogBinomialMod1Ggplot script from gwasanalytics package . This script is a modification of the bed2GwasCatalogBinomialGgplot and calculates binomial p-value for genomics overlaps using the following criteria . The P-values were computed using binomial cumulative distribution function b ( x;n , p ) in R ( dbinom function ) . We set the parameter n equal to the total number of GWAS SNPs in a particular GWAS phenotype . Parameter x was set to the number of GWAS SNPs for a given GWAS phenotype that overlap input regions and parameter p was set to the fraction of the uniquely mappable human hg19 genome ( calculated with subscript ) that is localized in the input regions and contains assessed GWAS phenotype SNPs . Calculated binomial p-value equals the probability of having x or more of the n test genomic regions in the open chromatin domain given that the probability of that occurring for a single GWAS genomic location is p . Plots were made using ggplot2 package and the wes anderson color palette in R ( https://github . com/karthik/wesanderson ) . All experiments were performed by the investigators blinded to the treatments/conditions during the data collection and analysis , using at least two independent preparations and treatments/conditions in triplicate . R/Bioconductor or GraphPad Prism 6 . 0 was used for statistical analysis . For enrichment analyses , we used both Fisher’s exact test and the cumulative binomial distribution test , as indicated . For comparisons between two groups of equal sample size ( and assuming equal variance ) , an unpaired two-tailed Student’s t-test was performed or in cases of unequal sample sizes or variance a Welch’s unequal variances t-test was performed , as indicated . P values <0 . 05 were considered statistically significant . For multiple comparison testing , one-way analysis of variance ( ANOVA ) accompanied by Tukey’s post hoc test were used as appropriate . All error bars represent standard error of the mean ( SE ) . The BiKE study is approved by the Ethical Committee of Northern Stockholm with following ethical permits: EPN DNr 95–276/277; DNr 02–146; DNr 02–147 , DNr 2005/83-31; DNR 2009/512-31/2; DNR 2009/295-31/2; 2011/950-32; 2012/619-32 and 213/2137-32 . The project is performed under the Swedish biobank regulations and prospective sampling is approved with informed consent procedure ( DNr 2009/512-31/2 ) . BiKE is registered at Socialstyrelsen ( The National Board of Health and Welfare ) and Biobank of Karolinska and approved by the Swedish Data Inspection Agency ( approval date/number 2002-09-30 DNr 916–2002 ) . All samples are collected with oral and written informed consent from patients or organ donor guardians . All animal procedures described in this study were approved by the Institutional Animal Care and Use Committees of Stanford University and conformed to NIH guidelines for care and use of laboratory animals . Specifically , the animal studies were approved by APLAC protocol #10022 , last approved on 3-16-17 and will remain in effect until 12-12-19 .
|
Coronary heart disease is the leading cause of death in the world . Both genes and the environment are important risk factors for the progression of disease , however , how genes may modulate the harmful response to the disease promoting environment is unknown and difficult to study . Here , we show that a common heritable variation in the gene TCF21 may regulate coronary heart disease risk by regulating the response of downstream gene activation by the disease environment . We find that a well-known environmental sensor , aryl-hydrocarbon receptor ( AHR ) , is regulated by TCF21 and also interacts with TCF21 , resulting in regulation of pro-inflammatory gene expression in coronary artery smooth muscle cells . We further show that oxidized LDL , a well-known driver of atherosclerosis in the plaque can activate the AHR pathway . This work describes a heritable form of gene-environment interaction identified through genome wide association studies in coronary artery disease , and presents an opportunity to define causal gene-gene and gene-environment interactions .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"genome-wide",
"association",
"studies",
"medicine",
"and",
"health",
"sciences",
"gene",
"regulation",
"regulatory",
"proteins",
"dna-binding",
"proteins",
"dna",
"transcription",
"coronary",
"heart",
"disease",
"genome",
"analysis",
"transcription",
"factors",
"cardiology",
"small",
"interfering",
"rnas",
"proteins",
"gene",
"expression",
"genetic",
"loci",
"biochemistry",
"rna",
"nucleic",
"acids",
"genetics",
"biology",
"and",
"life",
"sciences",
"genomics",
"non-coding",
"rna",
"vascular",
"medicine",
"computational",
"biology",
"human",
"genetics"
] |
2017
|
TCF21 and the environmental sensor aryl-hydrocarbon receptor cooperate to activate a pro-inflammatory gene expression program in coronary artery smooth muscle cells
|
The 2009 H1N1 influenza pandemic ( pH1N1 ) led to record sales of neuraminidase ( NA ) inhibitors , which has contributed significantly to the recent increase in oseltamivir-resistant viruses . Therefore , development and careful evaluation of novel NA inhibitors is of great interest . Recently , a highly potent NA inhibitor , laninamivir , has been approved for use in Japan . Laninamivir is effective using a single inhaled dose via its octanoate prodrug ( CS-8958 ) and has been demonstrated to be effective against oseltamivir-resistant NA in vitro . However , effectiveness of laninamivir octanoate prodrug against oseltamivir-resistant influenza infection in adults has not been demonstrated . NA is classified into 2 groups based upon phylogenetic analysis and it is becoming clear that each group has some distinct structural features . Recently , we found that pH1N1 N1 NA ( p09N1 ) is an atypical group 1 NA with some group 2-like features in its active site ( lack of a 150-cavity ) . Furthermore , it has been reported that certain oseltamivir-resistant substitutions in the NA active site are group 1 specific . In order to comprehensively evaluate the effectiveness of laninamivir , we utilized recombinant N5 ( typical group 1 ) , p09N1 ( atypical group 1 ) and N2 from the 1957 pandemic H2N2 ( p57N2 ) ( typical group 2 ) to carry out in vitro inhibition assays . We found that laninamivir and its octanoate prodrug display group specific preferences to different influenza NAs and provide the structural basis of their specific action based upon their novel complex crystal structures . Our results indicate that laninamivir and zanamivir are more effective against group 1 NA with a 150-cavity than group 2 NA with no 150-cavity . Furthermore , we have found that the laninamivir octanoate prodrug has a unique binding mode in p09N1 that is different from that of group 2 p57N2 , but with some similarities to NA-oseltamivir binding , which provides additional insight into group specific differences of oseltamivir binding and resistance .
The 2009 pandemic swine origin influenza A H1N1 virus ( pH1N1 ) has reminded the world of the threat of pandemic influenza [1] , [2] , [3] . In 2009 , the total sales of Tamiflu ( oseltamivir phosphate ) increased to over 3 billion US dollars ( Annual General Meeting of Roche Holding Ltd , 2 March 2010 ) . The total sales of Relenza ( zanamivir ) in 2009 were over 1 billion ( GlaxoSmithKline Quarter 4 Report , 4 February 2010 ) . Additionally , 5 . 65 million packs of Tamiflu were donated to the WHO in 2009 to replenish their stockpiles ( Roche , Annual General Meeting of Roche Holding Ltd , 2 March 2010 ) . Since the WHO has downgraded the threat of pH1N1 from the pandemic level in August 2010 , there have still been ongoing reports of pH1N1 outbreaks in south-eastern states of the USA , India and New Zealand ( US CDC ) . Furthermore , a new variant of pH1N1 has even been detected in Singapore , New Zealand and Australia [4] . Throughout the world , vaccinations have still been strongly advocated and stockpiles of oseltamivir and zanamivir are on reserve in case of another severe influenza outbreak in the near future . Both oseltamivir and zanamivir are excellent examples of modern structure-based drug-design and function as competitive inhibitors of the influenza neuraminidase ( NA ) , and are by far the most commonly used influenza drugs [5] , [6] , [7] , [8] . Influenza A virus contains two proteins on its surface in addition to the ion channel M2: hemagglutinin ( HA ) and NA [9] . Both M2 and NA are targets for clinically-available influenza drugs , however M2 drugs are rarely used anymore because M2 develops drug-resistant mutations very easily [10] . In the influenza virus infection life cycle , HA binds to terminally linked sialic acid receptors on the surface of host cells , allowing the virus to gain entry . In order for the influenza virus to efficiently break free from already infected cells and to continue replicating , sialic acid containing HA receptors must be destroyed . NA , which is a sialidase , catalyzes hydrolysis of terminally linked sialic acid and functions as the receptor-destroying element of influenza A and B viruses . Influenza A NA has been grouped into 9 different serotypes , N1-N9 , based upon antigenicity [11] . Additionally , influenza A NA is further classified into two groups: group 1 ( N1 , N4 , N5 and N8 ) and group 2 ( N2 , N3 , N6 , N7 and N9 ) , based upon primary sequence [12] . Group 1 NAs contain a 150-cavity ( formed by amino acids 147–151 of the 150-loop ) in their active site , whereas group 2 NAs lack this cavity [12] . Coordination of the 150-loop with the 430-loop appears to be critical for the formation of the 150-cavity [13] , [14] . Soaking experiments of typical group 1 NAs with inhibitors often result in the closure of the 150-cavity and indicates some flexibility of the 150-loop [12] , [15] . Molecular dynamics simulations also indicate some differences in the flexibility of the 150-loop between group 1 and group 2 NAs [14] , [16] . Structural studies reveal that Asp151 and Arg152 of the 150-loop form key interactions with the 4-group and N-acetyl group of common NA ligands , respectively . These two residues move away from the substrate in the open conformation of the 150-loop and closer upon ligand binding [17] . Therefore the 150-loop plays an essential role in substrate and inhibitor binding [15] . Furthermore , the 150-cavity is currently being successfully explored as a target for novel NA inhibitors [12] , [18] , [19] , [20] . The design of NA inhibitors is a classic example of structure-based drug discovery , pioneered by Mark von Itzstein and colleagues with the advent of the N2 , N9 and influenza B NA structures [5] , [6] , [8] , [21] , [22] , [23] , [24] . Currently there are four NA-targeting inhibitors that have been approved for use: zanamivir , oseltamivir , peramivir and laninamivir ( laninamivir has recently been approved in Japan ) . Additionally , there are many more NA inhibitors under clinical trials or under vigorous development due to the public threat of seasonal and pandemic flu and the rise of drug-resistant viruses [18] , [19] , [25] , [26] , [27] , [28] . However , previous results have indicated that inhibitors which are highly similar to the natural NA ligand , sialic acid , are less susceptible to the problem of drug-resistance [29] , [30] , [31] . This suggests that drugs like zanamivir , that are similar to sialic acid and its transition state analogue 2-deoxy-2 , 3-dehydro-N-acetylneuraminic acid ( Neu5Ac2en or DANA ) , have an advantage over oseltamivir , which is less similar ( Figure 1 ) . However , zanamivir must be administered twice daily over 5 consecutive days to attain its maximum effect . Therefore , the development of novel inhibitors that possess long term efficacy and that are also effective against oseltamivir-resistant influenza viruses is in great demand . Laninamivir ( R-125489 ) is a very promising , novel influenza NA inhibitor with high potency and the ability to efficiently inhibit common oseltamivir-resistant viruses , including those with the ubiquitous His274Tyr substitution [32] , [33] , [34] . Recently , laninamivir and its prodrug , laninamivir octanoate ( CS-8958 ) have been approved for use in Japan as Inavir ( Daiichi Sanko Press Release , 10 Sept . 2010 ) . Clinical studies have confirmed that the prodrug , laninamivir octanoate , is effective in both children and adults , however laninamivir octanoate has not yet been demonstrated to be more effective than oseltamivir against oseltamivir-resistant His274Tyr H1N1 infection in adult patients [35] , [36] , [37] . Like zanamivir , the core structure of laninamivir is Neu5Ac2en , the NA transition state analogue ( Figure 1 ) . Both laninamivir and zanamivir contain a 4-guanidino group that is not present in Neu5Ac2en and laninamivir also contains an additional 7-methoxy group ( Figure 1 ) . Laninamivir octanoate is the octanoyl prodrug of laninamivir ( Figure 1 ) . In a similar manner that oseltamivir is processed to oseltamivir carboxylate in the liver , it has been demonstrated that laninamivir octanoate is processed to laninamivir in the lung [33] . The laninamivir 7-methoxy and its prodrug octanoyl ester increase the ability to be retained in the lungs and to function effectively in a single inhaled dose [32] , [33] , [34] , [37] , [38] . Moreover , the high similarity of laninamivir to the NA transition state analogue , Neu5Ac2en , allows for an effective response against oseltamivir-resistant NA [32] , [33] , [34] , [35] . For these reasons , laninamivir and laninamivir octanoate offer advantages over both oseltamivir and zanamivir . In order to comprehensively assess the effectiveness of the novel NA inhibitors , laninamivir and laninamivir octanoate , in comparison to oseltamivir and zanamivir , and to reveal the structural basis of their inhibition , we utilized: 1 ) pandemic A/RI/5+/1957 H2N2 N2 ( p57N2 ) as a typical group 2 NA , 2 ) p09N1 as an atypical group 1 NA , and 3 ) avian H12N5 NA ( N5 ) as a typical group 1 NA . Soluble , active p57N2 , p09N1 and N5 were expressed in a baculovirus expression system and purified based upon previously reported methods [13] , [39] , [40] . NA inhibition assays were carried out and complex crystal structures were solved for laninamivir , laninamivir octanoate , zanamivir and oseltamivir in order to elucidate the structural basis of their inhibition . Our results indicate that laninamivir is potent against all 3 NAs with a similar binding mode to zanamivir . Laninamivir and zanamivir were more effective against group 1 N5 , with a 150-cavity , than atypical group 1 p09N1 and group 2 p57N2 , with no 150-cavity . This indicates that the ability of the bulky 4-guanidino group of zanamivir and laninamivir to become buried deep beneath the 150-loop is an important factor for their group-specific binding and inhibition . Furthermore , we confirm the binding of the prodrug , laninamivir octanoate , to p57N2 , with a similar binding mode to laninamivir . Surprisingly , the p09N1-laninamivir octanoate complex shows a completely different binding mode: p09N1 adopts a Glu276-Arg224 salt bridge in its laninamivir octanoate complex , forming a hydrophobic pocket that is also necessary to accommodate oseltamivir . The observation of different Glu276 rotation in p09N1 and p57N2 offers insight into the group specific differences of oseltamivir binding and resistance .
In our previous studies , we have successfully obtained both soluble p09N1 and N5 using a baculovirus expression system originally developed by Xu et al . [13] , [15] , [39] . To determine the functional and structural basis of NA inhibition and binding by laninamivir and its prodrug , we first expressed and purified a new group 2 member from the 1957 pandemic H2N2 virus , p57N2 , using similar methods . In this way , three major types of known NAs are covered in this comprehensive analysis: typical group 2 p57N2 , atypical group 1 p09N1 and typical group 1 N5 . p57N2 was crystallized and its structure was solved by molecular replacement using A/TOKYO/3/1967 ( H2N2 ) N2 ( PDB code: 1IVG ) as a search model [41] . As expected , the active site of p57N2 is highly similar to other group 2 NAs in that it has no 150-cavity ( Figure 2 - upper left ) . Like the available group 2 A/TOKYO/3/1967 ( H2N2 ) N2 and A/Memphis/31/98 ( H3N2 ) N2 structures [42] , [43] , p57N2 also contains a 150-cavity deficient active site with a salt bridge between Asp147 and His150 , confirming the presence of a stable , closed 150-loop ( Figure 2 - upper left; Figure 3A ) . Although the atypical group 1 p09N1 also has a 150-cavity deficient active site ( Figure 2 - upper right ) , the 150-loop is quite different from that of p57N2 . The p09N1 150-loop sequence ( residues 147–150 ) is GTIKD , however p57N2 contains DTVHD with 3 polymorphic amino acids . The p09N1 therefore contains no Asp147-His150 salt bridge , but instead contains Ile149 , which is commonly found in group 2 NAs , and Ile149 is able to rest closer to the hydrophobic Pro431 than Val149 is [13] . N5 on the other hand contains Val149 with no 147–150 salt bridge and displays a 150-cavity like all other structure-known NAs with Val149 and no 147–150 salt bridge ( Figure 2 - lower left ) [15] . Therefore , NAs with the three major styles of the 150-loop are covered in our comparative analysis . All NA proteins produced in the baculovirus expression system displayed stable sialidase activity . IC50 values and 95% confidence intervals ( CIs ) are given in Table 1 . Oseltamivir inhibited the activity of N5 , p09N1 and p57N2 with IC50 values of 0 . 83 nM , 0 . 54 nM and 0 . 79 nM , respectively . Laninamivir inhibition was best for group 1 N5 , followed by atypical group 1 p09N1 and worst for group 2 p57N2 . Zanamivir was also more effective against N5 than p09N1 and p57N2 , however the difference of zanamivir inhibition between p09N1 and p57N2 is not statistically significant ( Table 1 ) . Zanamivir inhibited N5 , p09N1 and p57N2 with IC50 values of 0 . 59 nM , 1 . 11 nM and 1 . 36 nM , respectively . Laninamivir was in a similar range with zanamivir for N5 , p09N1 and p57N2 with IC50 values of 0 . 90 nM , 1 . 83 nM and 3 . 12 nM , respectively . However inhibition of laninamivir was 1 . 53 , 1 . 65 and 2 . 29 fold lower than zanamivir for N5 , p09N1 and p57N2 , respectively . Inhibition of N5 , p09N1 and p57N2 by laninamivir octanoate was not as efficient , with IC50 values of 389 nM , 947 nM , and 129 nM , respectively . Hence inhibition of p57N2 by laninamivir octanoate was much better than for p09N1 . To determine the structural basis of the inhibition of laninamivir relative to zanamivir , we solved the very first complex structures of laninamivir with p57N2 , p09N1 and N5 at resolutions of 1 . 8 Å , 1 . 8 Å and 1 . 6 Å , respectively; and the complex structures of zanamivir with p57N2 , p09N1 and N5 at 1 . 9 Å , 1 . 9 Å and 1 . 6 Å , respectively [15] . Like zanamivir , the binding mode of laninamivir to all 3 NAs is highly similar to that of the NA transition state analogue , Neu5Ac2en . Some minor differences in the NA-inhibitor interactions between laninamivir and zanamivir are observed within each of the 3 NAs due to the additional hydrophobic 7-methoxy group of laninamivir; however all of the laninamivir complex structures highly resemble zanamivir binding ( Table 2 , Figure 3 ) . Due to the similar binding modes of zanamivir and laninamivr , we first carried out a detailed analysis of interactions with the 150-loop in each inhibitor complex . In all of the zanamivir and laninamivir structures , the 4-guanidino group is buried deep beneath the 150-loop where it forms many key hydrogen bonds with Glu119 , the Trp178 peptide carbonyl , Glu227 , and the Asp151 side chain and peptide carbonyl ( Figure 3 , Table 2 ) . This 4-guanidino group is the most buried part of the inhibitor in the structure ( Figure 3 ) , which is emphasized by the absence of any water molecules beneath the 150-loop and surrounding the 4-guanidino group . Although the 4-guanidino plays an essential role for the high affinity of laninamivir and zanamivir to NA , accessibility of the 4-guanidino to its binding site deep below the 150-loop is a crucial factor for the laninamivir and zanamivir binding process . The typical group 1 N5 contains a 150-cavity in its uncomplexed structure and inhibition of N5 by laninamivir and zanamivir was better than inhibition of p09N1 and p57N2 , which contain no 150-cavity in their uncomplexed structures ( Figure 3 , Table 1 ) . Therefore , our data indicate that the group specific accessibility of the laninamivir and zanamivir 4-guanidino to the NA active site is a key factor in determining their effectiveness . Slight differences in the interactions between the binding of laninamivir and zanamivir were observed due to the additional laninamivir 7-methoxy group ( Table 2 ) . Although this laninamivir 7-methoxy group is oriented away from its own ring oxygen and is pointed toward the hydrophobic Ile222 side chain , its distance is relative far at over 5 Å . Interactions between Arg371 and the inhibitor carboxylate were always highly consistent; however the carboxylate-Arg118 interactions are closer in zanamivir than laninamivir in every NA complex ( Table 2 ) . On the other hand , the carboxylate-Arg292 interactions are further in zanamivir than laninamivir in every NA complex ( Table 2 ) . Unlike p09N1 and p57N2 , N5 contains Tyr347 , which forms an additional hydrogen bond with the carboxylate of each inhibitor ( Figure 3C ) . Laninamivir octanoate complex structures with p09N1 and p57N2 ( Figure 4 ) were solved at 1 . 6 Å and 2 . 2 Å , respectively , demonstrating that the laninamivir octanoate prodrug can also directly inhibit NA without further processing . In p57N2 , laninamivir octanoate binds in a similar manner to laninamivir with an additional , novel hydrogen bond between the 9-ester carbonyl and Arg224 ( Figure 4A ) . p09N1 , on the other hand , has a totally different binding mode where the prodrug's ester is oriented toward Asn294 rather than Arg224 ( Figure 4B ) . p09N1 Glu276 is also in a different orientation in the laninamivir octanoate complex structure than in the zanamivir or laninamivir complex structures and forms a salt bridge with Arg224 in the same manner as oseltamivir binding ( Figure 4 and 5 ) . The rotation of p09N1 Glu276 places it out of range for hydrogen bonding with the 8-OH and 9-ester-O of laninamivir octanoate . Instead , the p09N1-laninamivir octanoate 9-ester-O forms a unique hydrogen bond with Asn294 ( Figure 4B ) . Additionally , 09N1 Ser247 forms another hydrogen bond with the laninamivir octanoate 9-ester-O at 3 . 4–3 . 5 Å . Still , the Glu276 rotation results in less hydrogen bonding in the p09N1-laninamivir octanoate complex compared to p57N2 ( Figure 4C , Table 2 ) . In both structures , there is no observed electron density corresponding to the octanoyl carbon chain indicating that this part of the molecule is highly flexible and does not engage many stable hydrophobic interactions with p09N1 or p57N2 . Still , electron density surrounding the entire ester can be observed in both complex structures . Furthermore , in p09N1 , the position 7-methoxy of laninamivir octanoate is also oriented slightly away from its N-acetyl group relative to laninamivir and there is additional electron density pointing toward the ring , indicating lower stability of the p09N1-laninamivir octanoate complex ( Figure 4 ) . In all of our NA complex structures , bond distances in each molecule of the asymmetric units are very similar , however in the p09N1-laninamivir octanoate structure some greater differences are observed between molecule A and B in the asymmetric unit , which further reflects the lower stability of the prodrug's octanoyl ester in p09N1 . In p09N1 molecule A , the distance between the 9-ester-O and Asn294 is 2 . 64 Å , however in molecule B the distance is much further at 3 . 93 Å ( Table 2 ) . To our surprise , the binding mode of the p09N1-laninamivir octanoate complex structure is similar to all known NA-oseltamivir complex structures with respect to the Glu276-Arg224 interactions . Therefore , we also solved the p09N1 oseltamivir complex structure at a resolution of 1 . 7 Å . As observed in the other available oseltamivir-NA complex structures , in the p09N1-laninamivir octanoate complex structure , Glu276 indeed forms a salt bridge with Arg224 , creating a hydrophobic pocket which accommodates the hydrophobic oseltamivir pentyl ether group ( Figure 5 ) [7] , [29] , [44] , [45] . This hydrophobic side chain of oseltamivir is favorably parallel to the Cβ and Cγ of Glu276 on one end and at the other end is pointed toward the hydrophobic Ile222 side chain , which contributes significantly to the high level of oseltamivir inhibition .
Recent studies have demonstrated that the novel influenza A virus NA inhibitors , laninamivir and laninamivir octanoate , are highly effective and have some advantages over zanamivir and oseltamivir [33] , [34] , [35] , [37] , [38] . In this study we verify that laninamivir , which highly resembles the NA transition state analogue , Neu5Ac2en , is indeed effective at inhibiting highly purified p57N2 , p09N1 and N5 , representing the three major types of all structure-known NAs with distinct 150-loop properties . Zanamivir and laninamivir are clearly more similar to sialic acid and Neu5Ac2en , than oseltamivir , which renders zanamivir and laninamivir less susceptible to drug-resistance and effective against many oseltamivir-resistant viruses [29] , [30] , [31] . The high degree of similarity between the binding modes of zanamivir and laninamivir in all of the NA complex structures indicates that zanamivir and laninamivir should be effective against the same drug-resistant mutations . However , laninamivir contains an additional 7-methoxy group which is oriented toward Ile222 . Although the distance between the laninamivir 7-methoxy and Ile222 is relative far ( over 5 Å ) , laninamivir may be susceptible to Ile222Arg , a rare drug-resistant substitution [46] . Moreover , the additional 7-methoxy group of laninamivir may disrupt hydrogen bonding of the 7-O with water and likely contributes to a slightly lower inhibition of laninamivir compared to zanamivir that was observed for all 3 NAs here ( Table 1 ) . Although zanamivir and laninamivir are highly similar to Neu5Ac2en , they both contain an artificial bulky 4-guanidino group . Upon binding , this 4-guanidino group becomes buried deep beneath Asp151 of the closed 150-loop and forms many hydrogen bonds which contribute to the high affinity of zanamivir and laninamivir to NA . However , the bulky 4-guanidino must be able to clear the 150-loop in order to bind NA and therefore a closed 150-loop may hinder the entry of zanamivir and laninamivir into the NA active site . In the open state of the 150-loop , when the 150-cavity is formed , Asp151 shifts over 1 . 5 Å ( the Asp151 Cγ is shifted over 2 Å in N5 ) away from the ligand binding site , which should facilitate entry of inhibitors like zanamivir and laninamivir [12] , [15] , [17] . A similar model has recently been proposed by Wang et al . , however this was based on a computer simulation using only the group 1 H5N1 NA structure [47] . The group specific 150-loop accessibility , based upon our structures of p57N2 , p09N1 , and N5 , is consistent with the inhibition efficiency of laninamivir and zanamivir . Group 2 p57N2 contains an Asp147-His150 salt bridge , limiting the flexibility of its closed 150-loop and inhibition of p57N2 by laninamivir was the lowest ( Figure 3A ) . p09N1 is an atypical group 1 with no 150-cavity , but no Asp147-His150 salt bridge , and inhibition of p09N1 by laninamivir was better than p57N2 ( Figure 3B ) . The typical group 1 N5 contains a 150-cavity in its uncomplexed structure and inhibition of N5 by both laninamivir and zanamivir was the highest ( Figure 3C ) . Therefore , we provide structural and functional evidence that the open 150-loop of a typical group 1 NA may facilitate the entry of the 4-guanidino group of zanamivir and laninamivir into the NA active site , relative to the closed 150-loop of group 2 NAs . The additional hydrogen bond between Tyr347 and the inhibitor carboxylate is also a key factor in explaining the higher N5 inhibition relative to p09N1 and p57N2 . However , like the closed 150-loop , this residue also makes the active site cavity smaller and in this way may also limit access of inhibitors to the N5 active site . Furthermore , this residue is found only in group 1 NAs which contain an open 150-loop cavity [12] . Thus , Tyr347 may compensate for the open 150-loop in regards to substrate binding . The complex structure of p57N2 with the laninamivir octanoate prodrug has a similar binding mode to laninamivir and zanamivir , however laninamivir octanoate in complex with p09N1 is completely different . This is the first time , as far as we know , that the same inhibitor has been observed to bind in two completely different conformations to influenza NAs . Additionally , p57N2 Arg224 forms a unique hydrogen bond with the laninamivir octanoate 9-ester carbonyl , and p09N1 Asn294 and Ser247 form unique hydrogen bonds with the laninamivir octanoate 9-ester-O . However , the novel conformation of the laninamivir octanoate-p09N1 complex disrupts any hydrogen bonding with Glu276 . The overall lack of hydrogen bonds and instability in the p09N1-laninamivir octanoate structure relative to p57N2 provides the structural basis for higher laninamivir octanoate inhibition of p57N2 observed in our study and a previous report demonstrating better laninamivir octanoate inhibition of H2N2 and H3N2 viruses over H1N1 viruses [34] . The absence of any electron density surrounding the octanoyl carbon chain of laninamivir octanoate indicates that it is unable to take part in any favorable interactions with p57N2 and p09N1 . The disordered octanoyl carbon chain likely destabilizes the interactions between the NA active site and the laninamivir octanoate 8-OH and 9-ester , which is indicated by the lower electron density surrounding the 9-ester . Therefore , the lower inhibition efficiency of laninamivir octanoate relative to laninamivir is not surprising . Binding of oseltamivir to p09N1 was indeed highly similar to the binding observed in previous reports and is also similar to the binding mode of laninamivir octanoate to p09N1 . Oseltamivir contains a 4-amino group , instead of the 4-guanidino group found in zanamivir and laninamivir , and is actually more similar to the natural ligand in this regard . Therefore , the orientation of the 150-loop during oseltamivir binding is not a major factor . Instead , the binding preference of oseltamivir for p09N1 over p57N2 and N5 may be instead explained by the ability of Glu276 to adopt the conformation that is critical to accommodate the osetalmivir pentyl ether side chain , which replaces the glycerol moiety of zanamivir , laninamivir and sialic acid . The observation that this Glu276 conformation occurs in the p09N1-laninamivir octanoate complex , but not the p57N2-laninamivir octanoate complex may indicate that this conformation is more stable in p09N1 after ligand binding which may explain why inhibition of p09N1 by oseltamivir was the best relative to N5 and p57N2 . In addition , this observation of different Glu276 dynamics in group 1 p09N1 compared to group 2 p57N2 offers some new insights into the group specificity of the oseltamivir-resistant His274Tyr substitution . The His274Tyr mutation is easily selected for N1 viruses , however cannot be selected for N2 virus types as N2 His274Tyr binding to oseltamivir is not impaired [48] . In group 2 NAs , Tyr274 is able to move away from Glu276 due to a small neighboring Thr252 residue , and oseltamivir can still bind for His274Tyr [30] . The native His274 is also further away from Glu276 in our p57N2-laninamivir octanoate structure and does not hydrogen bond with it . In group 1 NA , Tyr274 is not able to move away from Glu276 because of the bulky neighboring Tyr252 side chain , which prevents it from accommodating oseltamivir [30] . In a similar manner , the group 1 Tyr252 side chain promotes the native His274 to occupy a position where it can participate in a hydrogen bond network with Glu276 and Arg224 as observed in our 09N1-laninamivir octanoate structure ( Figure 4C ) . Recently , a clinical study has shown that laninamivir octanoate is not significantly better than oseltamivir against oseltamivir-resistant His274Tyr H1N1 infection in adult patients [36] . Since the laninamivir octanoate prodrug binds to p09N1 in a similar manner to oseltamivir , this may offer some explanation as to why laninamivir octanoate has a similar effect as oseltamivir against His274Tyr H1N1 . However , this may indicate that the laninamivir octanoate is not processed , or processed slowly , to laninamivir in the adult patients from this study , since laninamivir has been clearly demonstrated to be effective against the oseltamivir-resistant His274Tyr influenza A viruses [32] , [33] , [34] . Further investigation into the efficacy of laninamivir octanoate in adults in clearly needed . The results from this comprehensive analysis of group 2 p57N2 , atypical group 1 p09N1 and typical group 1 N5 support the hypothesis that influenza NA inhibitors which more closely resemble the NA transition state analogue , Neu5Ac2en , are more likely to remain effective against NAs from both groups and with various drug-resistant amino acid substitutions . Most importantly , we provide mechanisms to explain the group 1 preference of laninamivir and zanamivir and the differential binding of the octanoate prodrug to group 1 p09N1 and group 2 p57N2 derived from pandemic influenza viruses .
Methylumbelliferyl-N-acetylneuraminic acid ( MUNANA ) was purchased from J&K Scientific Ltd . Sialic acid ( Neu5Ac ) was purchased from Sigma ( Cat . No . 855650 ) and used without further purification . Laninamivir , laninamivir octanoate , zanamivir and oseltamivir were readily synthesized according to the relevant literatures [49] , [50] , [51] , [52] , [53] . All products were characterized by their NMR or MS spectra . NA was prepared in a baculovirus expression system according to methods based on an original method reported by Xu et al . [39] . Both N5 and p09N1 were prepared as previously described in our laboratory [13] , [15] . For p57N2 , the cDNA encoding amino acid residues 83–469 were recombined into the baculovirus transfer vector pFastBac1 ( Invitrogen ) , with a GP67 signal peptide , a 6X his-tag , a tetramerizing sequence and a thrombin cleavage site at the N-terminus . Recombinant baculovirus was prepared based on the manufacturer's protocol ( Invitrogen ) . Sf9 suspension cultures were grown in Sf-900 II SFM serum-free media ( GIBCO ) at 28°C and 120 RPM and transfected with high-titer recombinant baculovirus . After growth of the transfected Sf9 suspension cultures for 3 days , centrifuged media were applied to a HisTrap FF 5 mL column ( GE Health ) which was washed with 20–50 mM imidazole , and then NA was eluted using 200–300 mM imidazole . After dialysis , thrombin digestion ( Sigma , 3 U/mg NA; overnight at 4°C ) and gel filtration chromatography using a Superdex-200 10/300 GL column ( GE Healthcare ) , NA fractions were analyzed by SDS-PAGE . High-purity NA fractions were pooled and concentrated using a membrane concentrator with a molecular weight cutoff of 10 kD ( Millipore ) . A buffer of 20 mM Tris-HCl , 50 mM NaCl , pH 8 . 0 was used for gel filtration and protein concentration . A neuraminidase inhibition assay using MUNANA was performed as described by Potier et al . with modifications [54] . Briefly , 10 uL of purified , recombinant NA ( 10 nM ) was mixed with 10 uL of inhibitor and incubated for 30 min at room temperature . NA and inhibitors were carefully diluted in fresh PBS buffer . At least 5 concentrations of each inhibitor at an appropriate range were used for each repeat . Following incubation , 30 uL of 166 uM MUNANA in 33 mM MES and 4 mM CaCl2 ( pH 6 . 0 ) was added to the solution to start the reaction using a 12-tip pipette ( Eppendorf ) . A positive and a negative control were included in each 12-well lane . After starting the reaction for each lane on the plate , the reaction mixture was immediately loaded on a SpectraMax M5 ( Molecular Devices ) where fluorescence was quantified over the course of 30 min at an excitation wavelength of 355 nm and an emission wavelength of 460 nm . Single time points were chosen where the positive control produced a fluorescence signal of approximately 1 , 000 . All assays were done in triplicates and IC50 values for each inhibitor were calculated with sigmoidal fitting of the log[inhibitor] vs . inhibition percentage using GraphPad Prism . NA crystals were grown using the hanging-drop vapor diffusion method . Initial screening was performed using a commercial kit ( Hampton Research ) . Diffraction quality crystals of p57N2 were obtained by mixing 1 uL of the concentrated protein at 10 mg/mL in 20 mM Tris , pH 8 . 0 , and 50 mM NaCl with 0 . 1M BIS-TRIS propane ( pH 9 . 0 ) , 10% v/v Jeffamine ED-2001 ( pH 7 . 0 ) . N5 crystals were obtained using 0 . 1 M HEPES ( pH 7 . 5 ) , 12% w/v polyethylene glycol 3 , 350 at 18°C [15] . Quality p09N1 crystals were obtained as described previously using 0 . 16 M calcium acetate hydrate , 0 . 08 M sodium cacodylate trihydrate , pH 6 . 5 , 14 . 4% polyethylene glycol 8000 , 20% glycerol at 18°C [13] . NA protein crystals were first incubated in mother liquor containing 20 mM of inhibitor , and then flash-cooled at 100 K . Diffraction data for the p57N2 and N5 were collected at KEK beamline Ne3A , while p09N1 data were collected at SSRF beamline BL17U . Diffraction data were processed and scaled using HKL2000 [55] . Data collection and processing statistics are summarized in Table 3 . The structure of p57N2 was solved by molecular replacement method using Phaser [56] from the CCP4 program suite [57] with the structure of A/TOKYO/3/1967 H2N2 N2 ( PDB code: 1IVG ) as the search model [41] . Initial restrained rigid-body refinement and manual model building were performed using REFMAC5 [58] and COOT [59] , respectively . Further rounds of refinement were performed using the phenix . refine program implemented in the PHENIX package with coordinate refinement , isotropic ADP refinement and bulk solvent modeling [60] . The stereochemical quality of the final model was assessed with the program PROCHECK [61] . The final models have 84% of the residues in the most favored region of the Ramachandran plot [62] and no residue in disallowed regions . Structures of p09N1 and N5 were solved as described previously [13] , [15] . All crystal structures have been deposited into the Protein Data Bank ( PDB , www . pdb . org ) with the following PDB codes: N5-laninamivir - 3TI8 , p09N1-zanamivir - 3TI5 , p09N1-laninamivir - 3TI3 , p09N1-laninamivir octanoate - 3TI4 , p09N1-oseltamivir - 3TI6 , p57N2-zanamivir - 3TIC , p57N2-laninamivir - 3TIA , and p57N2-laninamivir octanoate - 3TIB .
|
The influenza neuraminidase ( NA ) enzyme is the most successful drug target against the seasonal and pandemic flu . The 2009 H1N1 flu pandemic led to record sales of the NA inhibitors oseltamivir ( Tamiflu ) and zanamivir ( Relenza ) . Recently , a new drug , laninamivir ( Inavir ) , has been approved for use in Japan can also be administered effectively using a single dose via its octanoate prodrug ( CS-8958 ) , however its effectiveness against oseltamivir-resistant influenza infection has not been demonstrated in clinical studies . In this study we comprehensively evaluate the effectiveness of laninamivir and its prodrug using NA from different groups with different active site features . We expressed and purified a group 2 NA from the 1957 pandemic H2N2 , an atypical group 1 NA from the 2009 H1N1 pandemic and a group 1 NA from avian H12N5 . NA inhibition was assayed and NAs were further crystallized with each inhibitor to determine the structural basis of their action . We found that laninamivir inhibition is highly potent for each NA , however binding and inhibition of laninamivir and its prodrug showed group specific preferences . Our results provide the structural and functional basis of NA inhibition using classical and novel inhibitors , with NAs from multiple serotypes with different properties .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biochemistry",
"proteins",
"protein",
"structure",
"biology",
"proteomics",
"glycobiology",
"drug",
"discovery"
] |
2011
|
Structural and Functional Analysis of Laninamivir and its Octanoate Prodrug Reveals Group Specific Mechanisms for Influenza NA Inhibition
|
Insulin-like signalling is a conserved mechanism that coordinates animal growth and metabolism with nutrient status . In Drosophila , insulin-producing median neurosecretory cells ( IPCs ) regulate larval growth by secreting insulin-like peptides ( dILPs ) in a diet-dependent manner . Previous studies have shown that nutrition affects dILP secretion through humoral signals derived from the fat body . Here we uncover a novel mechanism that operates cell autonomously in the IPCs to regulate dILP secretion . We observed that impairment of ribosome biogenesis specifically in the IPCs strongly inhibits dILP secretion , which consequently leads to reduced body size and a delay in larval development . This response is dependent on p53 , a known surveillance factor for ribosome biogenesis . A downstream effector of this growth inhibitory response is an atypical MAP kinase ERK7 ( ERK8/MAPK15 ) , which is upregulated in the IPCs following impaired ribosome biogenesis as well as starvation . We show that ERK7 is sufficient and essential to inhibit dILP secretion upon impaired ribosome biogenesis , and it acts epistatically to p53 . Moreover , we provide evidence that p53 and ERK7 contribute to the inhibition of dILP secretion upon starvation . Thus , we conclude that a cell autonomous ribosome surveillance response , which leads to upregulation of ERK7 , inhibits dILP secretion to impede tissue growth under limiting dietary conditions .
Animals coordinate tissue growth and body size with changing nutrient conditions and it is well established that insulin-like signalling has a key role in this process [1] , [2] . In mammals , insulin-like signalling is mediated by insulin and insulin-like growth factors through their respective receptors . Drosophila possesses a single insulin-like receptor , which is activated by insulin-like peptides ( dILPs ) . There are currently eight dILPs identified in Drosophila and several of them ( e . g . dILP2 , -3 and -5 ) are mainly expressed and secreted by a group of 14 median neurosecretory cells , also known as insulin-producing cells ( IPCs ) [3] , [4] . These cells are critical in controlling organismal growth , as disturbance of IPC function leads to strongly reduced body size [4] . IPC function is coupled to nutrient status as starvation of Drosophila larvae leads to inhibition of dILP secretion [5] . Recent studies have demonstrated that fat body , the insect counterpart of adipose tissue and liver , senses nutrient status and regulates dILP secretion by hormonal mechanisms [5] , [6] . Secreted fat body cytokine Unpaired 2 is regulated by dietary sugars and lipids and it controls a population of GABAergic neurons in the brain that project onto IPCs . Upd2 activates the JAK-STAT pathway in these neurons , which relieves the inhibitory effect of GABAergic neurons on IPCs , resulting in dILP secretion [6] . Moreover , the fat body senses amino acid levels and in turn secretes an unknown humoral signal , which facilitates dILP secretion from the IPCs [5] . Recently , it was shown that adiponectin receptor also regulates dILP secretion from the IPCs , but the corresponding ligand is not yet identified [7] . Beyond these examples , the regulation of IPC function in response to changing nutrient status is poorly understood . In particular , very little is known of how cell intrinsic pathways within the IPCs affect dILP secretion , and consequently organismal growth . Ribosome biogenesis is a major consumer of cellular energy and a key determinant of cell growth capacity [8] . Ribosome biogenesis is tightly regulated by nutrient , growth factor and stress responsive signalling networks [1] , [8]–[13] . As the ribosome biogenesis pathway is a major integrator of growth regulatory signals , its activity needs to be closely monitored . p53 serves as one of the ribosome biogenesis surveillance factors [14] . Impaired ribosome biogenesis activates p53 , likely through several parallel mechanisms . For example , imbalance of ribosomal components allows ribosomal proteins RPL5 and RPL11 in complex with 5S rRNA to inhibit HDM2/MDM2-dependent degradation of p53 [15]–[17] . On the other hand , nucleolar Myb binding protein 1A ( MYBBP1A ) translocates to the nucleoplasm if rRNA production is inhibited [18] . Nucleoplasmic MYBBP1A increases p53 activity by promoting its tetramerization and acetylation [19] . In proliferating cells , activation of p53 through these mechanisms leads to cell cycle arrest [16] . However , the role of ribosome biogenesis surveillance in other physiological settings is poorly understood . This is a relevant question , as disturbed ribosome biogenesis is known to have highly tissue-specific consequences in vivo . In humans , mutations in ribosomal components or other ribosome biogenesis genes lead to genetic disorders called ribosomopathies [20] . They manifest , for example , as bone marrow failure , anemia , skin abnormalities , and pancreatic insufficiency [20] . In Drosophila , mutations that disturb ribosome biogenesis lead to so-called Minute phenotype , displaying impaired growth , poor fertility and short and thin bristles [21] . Inhibition of ribosome biogenesis specifically in the Drosophila fat body or prothoracic gland influences systemic growth by affecting hormonal regulation through insulin and ecdysone signalling axes , respectively [22]–[24] . Which signalling mechanisms mediate the distinct cell type-specific phenotypes of ribosome surveillance in vivo is , however , poorly understood . Here we report that dILP secretion is tightly coupled to the activity of the ribosome biogenesis pathway within the IPCs . We show that the p53 dependent ribosomal surveillance response inhibits dILP secretion and identify atypical MAP kinase ERK7 as a downstream effector of the ribosome surveillance response in this setting . Furthermore , we provide evidence that both p53 and ERK7 activities within the IPCs contribute to regulation of dILP secretion in response to nutrient status . Thus , we propose that the p53- and ERK7-dependent ribosome surveillance pathway serves as a local branch of the IPC-regulating nutrient sensing network , parallel to the humoral signals derived from the fat body .
While the role of Drosophila insulin-like peptides ( dILPs ) in systemic growth control is well established , it remains poorly understood how the secretion of dILPs is regulated in response to changing dietary conditions . To get a better insight into this question , we explored the cell autonomous signalling mechanisms that regulate dILP secretion , by performing a kinome-wide screen with RNAi expressed specifically in the IPCs . Impaired IPC function compromises tissue growth [4] and therefore we used body weight of emerging adults as a screening readout ( Figure 1A , Table S1 ) . To normalize variation between vials due to growth conditions , we determined the relative weight wherein the mean weight of the flies expressing both dILP2-Gal4 and RNAi was divided by the mean weight of control flies ( not expressing the dILP2-Gal4 ) from the same vial . In total , 231 kinases were screened . All primary hits displaying >10% difference to the median relative weight were further analyzed by an independent RNAi line . Altogether 12 protein kinases showed significant body weight reduction in two independent RNAi lines ( Figure 1B ) . All identified kinases have mammalian orthologs ( Table 1 ) . In addition to these 12 kinases identified here , we have earlier shown that atypical PKC is essential for IPC function [25] . One of the kinase hits was TOR ( target of rapamycin ) . TOR kinase is present in two functionally distinct complexes , TOR complex 1 and 2 ( TORC1 and TORC2 ) . To explore which of the TOR complexes have a regulatory role in the IPCs , we depleted Raptor and Rictor , essential components of TORC1 and TORC2 , respectively . Knockdown of Raptor in the IPCs significantly reduced total body weight , while depletion of Rictor had no significant impact ( Figure 1C ) , showing that TORC1 regulates growth through the IPCs . However , as RNAi does not completely silence gene expression , we cannot rule out the involvement of TORC2 . To explore gene expression of three IPC-expressed dilp genes , namely dilp2 , dilp3 and dilp5 , we analyzed RNA from larval brains by quantitative RT-PCR ( Figure 1D ) . In the case of most hits , expression of dilp2 and dilp5 remained nearly unchanged . In contrast , dilp3 expression was elevated in several samples . This is likely due to a feedback mechanism activated by reduced autocrine insulin signalling in the IPCs [26] . Only depletion of Tousled-like kinase ( Tlk ) significantly reduced dilp3 expression ( Figure 1D ) . Secretion of dILPs is a key regulatory level in determining the activity of systemic insulin signalling [5] . dILP2 contributes to the total body weight [26] and its secretion can be assessed by monitoring its accumulation into the cell bodies of IPCs . Accumulation is observed when dILP2 secretion is inhibited upon starvation [5]; ( Figure 2A , B ) . Of the hits , 5 kinases caused significant dILP2 accumulation upon knockdown . These include Cdk12 , Adck ( CG3608 ) , Pkc98E ( Figure S1 ) , as well as two Rio kinases , Rio1 and Rio2 ( Figure 2A , B ) . Rio1 and Rio2 belong to a group of atypical kinases and they have a conserved role in ribosome maturation [27]–[30] . As ribosome biogenesis is a process tightly coupled to nutrient sensing [1] , [12] , we chose to further explore the role of Rio kinases in dILP secretion . We confirmed that , similarly to starvation , IPC-specific depletion of Rio kinases led to elevated insulin-like receptor ( inr ) gene expression in the larva ( Figure 2C ) , which is an established readout for reduced peripheral insulin signalling [31] . If the inhibition of dILP2 secretion was due to impaired ribosome biogenesis , we predicted that depletion of other ribosomal genes would phenocopy this effect . Indeed , IPC-specific knockdown of a number of other ribosomal components or genes involved in various steps of ribosome biogenesis led to significantly reduced body size ( Figure 2D ) . To further confirm that the dILP2 secretion phenotypes of Rio kinases were due to impaired ribosome biogenesis , we depleted ribosomal protein Rpl35A and ribosome assembly factor Tsr1 ( CG7338 ) in the IPCs and observed a similar inhibition of dILP2 secretion in both cases ( Figure 2E , F ) . The dILP2-Gal4 driver we used [4] displays activity in the salivary glands , in addition to IPCs . Depletion of Rio2 by a salivary gland specific Sgs3-Gal4 driver [32] caused no growth reduction , ruling out the possibility of nonspecific effects through the salivary glands ( Figure S2 ) . TORC1 signalling regulates ribosome biogenesis gene expression in Drosophila in a context-dependent manner . In S2 cells , expression of ribosome biogenesis genes is TORC1-dependent [33] , [34] , while in the context of whole larvae , TOR mutants do not display significantly reduced expression of ribosome biogenesis genes [35] . In accordance with the earlier observations in whole larvae , inhibition of Raptor in the IPCs did not lead to inhibited dILP2 secretion ( Figure S3 ) . Transcription factor dMyc is a conserved master regulator of ribosome biogenesis [13] , [36] . It transcriptionally regulates the expression of rRNAs as well as ribosomal proteins [36] . In Drosophila larvae ribosome biogenesis genes that are inhibited upon starvation are targets of dMyc [35] . Therefore we tested whether dMyc activity in the IPCs has an impact on dILP2 secretion . Indeed , RNAi-mediated depletion of dMyc prominently inhibited dILP2 secretion ( Figure 2G , H ) and led to reduced body weight ( Figure 2I ) . We wanted to further explore whether inhibition of ribosome biogenesis affects the secretion of other dILPs in the IPCs . Similarly to dILP2 , secretion of dILP5 was inhibited upon Rio2 knockdown ( Figure 2J , K ) . Inhibition of systemic growth often manifests as prolonged larval development , in addition to reduced body size . Indeed , depletion of Rio2 in the IPCs led to significant delay in pupation kinetics ( Figure 2L ) . In conclusion , inhibition of ribosome biogenesis in the IPCs triggers a response that blocks dILP secretion and consequently leads to slower larval development and reduced body size . How is the ribosome biogenesis pathway coupled to dILP secretion ? We hypothesised that one of the ribosome biogenesis surveillance pathways might link secretion to the status of ribosome biogenesis in specialised secretory cells like the IPCs . As p53 is the best-established surveillance factor for the ribosome biogenesis pathway [14] , [15] , [37] , we wanted to explore , whether it is involved in the regulation of dILP secretion . Indeed , overexpression of p53 in the IPCs revealed that p53 is sufficient to cause dILP2 accumulation ( Figure 3A , B ) . To analyse the possible impact of transcriptional regulation , we measured dilp2 mRNA levels by quantitative RT-PCR . p53 overexpression caused modest downregulation of dilp2 mRNA levels ( Figure S4 ) , ruling out the possibility of transcriptional activation as a cause for p53-dependent dILP2 accumulation . To assess the contribution of p53-mediated dILP regulation on growth , we analyzed pupal volume , which is a sensitive means to measure changes in total body size [22] . In our hands , pupal volume data is consistent with the adult weight , but displays less random variation . In accordance with the immunofluorescence data suggesting inhibited dILP2 secretion , p53 overexpression in the IPCs strongly reduced the pupal volume ( Figure 3C ) . In contrast , overexpression of p53 in the salivary glands caused no growth impairment ( Figure S2 ) . In sum , our data is consistent with the idea that p53 expression in the IPCs is sufficient to prevent dILP secretion and consequently inhibit tissue growth . Next , we wanted to explore , whether the p53 would mediate the observed inhibition of dILP2 secretion upon disturbed ribosome biogenesis . This was the case , as depletion of p53 rescued the dILP2 accumulation upon Rio2 knockdown ( Figure 3D , E ) along with partial rescue of the growth impairment ( Figure 3F ) . A similar rescue of growth was observed when p53 was depleted in combination with Rio1 ( Figure S5 ) . Moreover , a control RNAi targeting firefly Luciferase , but no endogenous genes , had no influence on the phenotypes of Rio2 RNAi in the IPCs ( Figure S6 ) , demonstrating that expression of an additional RNAi does not interfere with Rio2 RNAi activity . Identification of p53 as an essential inhibitor of dILP2 secretion implied that ribosome biogenesis regulates dILP secretion through an active surveillance mechanism , rather than through an indirect mechanism following reduced translation . To directly couple ribosome surveillance to secretion we turned our attention to known regulators of the secretory pathway . It was recently shown that an atypical MAP kinase called ERK7 ( also known as ERK8 or MAPK15 ) inhibits secretion upon serum and amino acid starvation in Drosophila S2 cells [38] . ERK7 was shown to phosphorylate Sec16 and thus prevent its recruitment to ER exit sites , consequently preventing the export of the secretory cargo . Therefore , we wanted to explore , whether ERK7 is regulated in response to inhibited ribosome biogenesis . In order to do so , we used in situ hybridization of larval brain . erk7 expression was undetectable in fed control larvae , whereas overexpression of transgenic ERK7 by dILP2-Gal4 led to a strong IPC-specific signal ( Figure S7 ) . As a negative control we used an erk7 sense probe for the ERK7 overexpressing samples ( Figure S7 ) . After confirming the specificity of the in situ hybridization , we analysed erk7 expression upon disturbed ribosome biogenesis . Intriguingly , we observed that erk7 mRNA levels were clearly elevated in the IPCs following Rio2 depletion using dILP2-Gal4 ( Figure 4A ) , implying that ERK7 is upregulated upon ribosome surveillance in the IPCs . As ribosome biogenesis is inhibited upon starvation [35] , we hypothesized that starvation might also cause elevated expression of erk7 , which indeed was the case ( Figure 4A ) . Notably , in the context of brain tissue , the starvation-induced upregulation appeared specific to the IPCs , since erk7 mRNA remained undetectable in other brain areas . To confirm the regulation of erk7 mRNA by an independent method , we used quantitative RT-PCR on mRNA samples isolated from whole larvae . Disturbance of ribosome biogenesis by ubiquitous depletion of Rio1 , Rio2 and Myc led to elevated erk7 expression ( Figure 4B , C ) . Similarly , starvation caused significant upregulation of erk7 mRNA levels in the whole larval samples ( Figure 4D ) . In sum , our data shows that erk7 gene expression is elevated upon impaired ribosome biogenesis and starvation , suggesting a possible role for ERK7 in the regulation of dILP secretion during these conditions . To this end , we decided to explore the functional relationship between ribosome surveillance , ERK7 and dILP secretion in the IPCs . We analyzed transgenic flies that specifically overexpress ERK7 in the IPCs ( Figure S7 ) . Indeed , ERK7 overexpression led to prominent accumulation of dILP2 and dILP5 in the IPCs ( Figure 5A–5D ) . Consistently , ERK7 overexpression in the IPCs , but not in salivary glands , led to reduced pupal volume ( Figure 5E and Figure S2 ) as well as delayed pupation ( Figure 5F ) . To rule out the possibility that the ERK7-dependent dILP accumulation was due to elevated transcription , we used quantitative RT-PCR to measure dilp2 and dilp5 mRNA levels , which remained unchanged upon ERK7 overexpression ( Figure S8 ) . Thus , we conclude that elevated ERK7 expression is sufficient to inhibit dILP secretion cell autonomously in the IPCs . We also performed genetic epistasis experiments to test if ERK7 is essential to inhibit dILP2 secretion upon disturbed ribosome biogenesis . Indeed , ERK7 RNAi efficiently suppressed the dILP2 accumulation caused by Rio2 knockdown ( Figure 5G , H ) . Similar suppression was observed in the case of dILP2 accumulation following dMyc depletion ( Figure S9 ) . Consistent with the effects on dILP2 secretion , IPC-specific depletion of ERK7 partially suppressed the impaired growth upon knockdown of Rio2 , while having no effect on body size in the wild type background ( Figure 5I ) . As we had observed that both p53 and ERK7 were essential for the inhibition of dILP2 secretion upon impaired ribosome biogenesis , we wanted to explore whether they belong to the same pathway . We hypothesized that ERK7 might be a downstream effector of p53 . Supporting this idea , we observed that overexpression of wild type p53 in the IPCs led to elevated expression of erk7 mRNA as detected by in situ hybridization ( Figure 6A ) . To further test the possible relevance of the p53-mediated regulation of erk7 in the IPCs , we performed genetic epistasis experiments . Indeed , depletion of ERK7 efficiently suppressed the dILP2 accumulation upon p53 overexpression ( Figure 6B , C ) . Consistently , ERK7 RNAi also partially suppressed the reduction of pupal volume caused by p53 overexpression ( Figure 6D ) along with the delay in pupation ( Figure 6E ) . Notably , the suppression by ERK7 RNAi was only partial , leaving open the possibility for a parallel , ERK7-independent , mechanism . In sum , our data implies that p53 and ERK7 belong to the same pathway and ERK7 acts as a downstream effector of p53 in inhibiting dILP2 secretion . As ribosome biogenesis is dependent on nutrients [1] , [12] , [35] , and erk7 expression was elevated upon starvation ( Figure 4A , D ) , we postulated that perhaps the p53- and ERK7-dependent ribosome surveillance response contributes to dILP accumulation in the IPCs upon starvation . This proved to be the case , as inhibition of either p53 ( Figure 7A , B; Figure S10 ) or ERK7 ( Figure 7C , D ) in the IPCs partially reduced the accumulation of dILP2 following starvation of the larvae . While we cannot rule out the possibility that the knockdown of p53 and ERK7 affects dILP2 translation in this setting , the most parsimonious explanation is that p53 and ERK7 contribute to dILP2 secretion upon starvation .
Here we report a novel cell-autonomous control mechanism for dILP secretion . Specifically , we conclude that i ) inhibition of ribosome biogenesis in the IPCs at any level tested , including ribosomal gene expression ( Myc ) , ribosome maturation ( Rio1 , Rio2 , Tsr1 ) or by introducing imbalance of ribosomal components ( Rpl35A ) , triggers a response to inhibit dILP secretion , ii ) this inhibitory response is dependent on p53 , a known surveillance factor for ribosome biogenesis , iii ) a downstream effector of this ribosome surveillance pathway is protein kinase ERK7 , as erk7 mRNA levels are elevated upon inhibited ribosome biogenesis and p53 activation and erk7 is essential to inhibit dILP secretion in both conditions , iv ) the ribosome surveillance mechanism discovered here likely contributes to starvation-induced inhibition of dILP secretion . These findings significantly broaden the view about the regulatory functions of the ribosome surveillance pathways , which have been mainly explored at the level of proliferating cells . Ribosome biogenesis serves as the key determinant of cell autonomous growth control and it is finely tuned to match with the cellular nutrient and energy status [1] , [12] . Coupling dILP secretion to the ribosome biogenesis pathway is an elegant mechanism for multicellular animals to synchronize the hormonal growth control with the local cell autonomous regulation of growth . Comparable to our finding in the IPCs , inhibition of ribosome biogenesis in the fat body leads to a block of dILP secretion in the IPCs through an unknown humoral mechanism [22] , [24] , [39] . Linking ribosome biogenesis to growth control through parallel mechanisms likely provides a robust regulatory network to tune down systemic growth signals whenever any region of the body experiences nutrient deprivation . This synchronization is likely important to maintain balanced growth throughout the spectrum of dietary conditions . Future studies should be aimed to explore the possible interrelationship between the fat body-derived signals and the cell-autonomous mechanism discovered here . Our findings highlight that the p53-mediated ribosome surveillance pathway can serve highly cell type-specific functions in vivo . This is interesting when considering human ribosomopathies , genetic diseases caused by impaired ribosome biogenesis manifesting with a wide spectrum of tissue-specific defects . One of the ribosomopathies , Shwachman-Diamond syndrome ( SDS ) , is manifested with a failure in pancreatic function [20] . A mouse model of SDS displays general preservation of ductal and endocrine compartments , but reduced amount of zymogen granules . Moreover , SDS mutant mice have reduced glucose tolerance , suggesting compromised endocrine function [40] . It will be interesting to learn whether p53 and ERK7 act as mediators of the secretion-related defects observed in SDS . Compared to other members of the MAP kinase family , ERK7 has remained relatively poorly characterized [41] . Earlier studies in mammalian and Drosophila cells have shown that ERK7 protein levels are actively regulated at the level of protein degradation [38] , [42] . In mammals , an increase in ERK7 levels leads to autophosphorylation and consequent activation [43] , [44] . Consistent with the idea that ERK7 is mainly regulated through abundance , we observed that elevated ERK7 expression had a prominent impact in the function of IPCs . Interestingly , earlier studies have linked ERK7 function to growth regulation by showing that ERK7 protein is stabilized by serum and amino acid starvation [38] . Our data provides evidence that impaired ribosome biogenesis as well as starvation increases the expression of erk7 mRNA revealing a novel regulatory level for ERK7 function . In addition to the conditions explored here , ERK7 expression levels increase towards the end of larval development when growth is ceased [45] . It will be interesting to learn further how ERK7 expression is regulated and whether ERK7 has a function in tissue growth control beyond its role in the IPCs . Earlier studies in cell culture have shown evidence that ERK7 regulates cancer cell proliferation and autophagy [46] , [47] , suggesting that ERK7 may have a broader role in the regulation of tissue growth .
RNAi lines were from Vienna Drosophila RNAi Center ( VDRC ) , Bloomington Drosophila Stock Center ( BDSC ) and National Institute of Genetics ( NIG ) , Japan . Identities of RNAi lines and transgenes are listed in Table S2 . Driver lines used: dILP2-Gal4 , UAS-GFP [4] for all IPC specific experiments , Tub-G80TS; Tub-Gal4 ( temperature sensitive ubiquitous driver ) for ubiquitous expression of RNAi . Sgs3-Gal4 ( BDSC#6870 ) was used for salivary gland specific experiments [32] . w1118 crossed to respective driver lines was used as control in all experiments . For generating UAS-ERK7-V5-His , ERK7 cDNA was amplified by PCR from pMT-ERK7-V5-HisB [38] , cloned into KpnI site of the pUAST-attB vector and confirmed by sequencing . The transgene was directed to the attP landing site at 22A2 . Transgenic flies were constructed by BestGene Inc . Flies were grown on medium containing agar 0 . 6% ( w/v ) , semolina 3 . 2% ( w/v ) , malt 6 . 5% ( w/v ) , dry baker's yeast 1 . 8% ( w/v ) , propionic acid 0 . 7% ( v/v ) and Nipagin ( methylparaben ) 2 . 4% ( v/v ) . For starvation , third instar larvae were placed on PBS , 1% sucrose , 0 . 4% agar for approximately 15 hours . To normalize variation between vials due to growth conditions , we determined the relative weight . UAS-RNAi flies were crossed to dILP2-Gal4 , UAS-GFP/CyO flies and progeny were allowed to develop at 29°C . Four days after adult emergence , male flies were weighed in groups ( ≥10 flies/group , N≥3 ) , using a precision balance ( Mettler ) . The mean weight of the flies expressing both dILP2-Gal4 and RNAi was divided by the mean weight of control flies ( not expressing the dILP2-GAL4 ) from the same vial . UAS-RNAi flies were crossed to dILP2-Gal4 , UAS-GFP/Cyo flies and allowed to lay eggs at +25°C . GFP positive L1 larvae were collected ( 30 per vial ) 24 hours after egg deposition ( AED ) and grown at +29°C until pupation . Length and width of pupae were measured using ProgRes CapturePro 2 . 8 . 8 program ( Jenoptik ) . Volumes of the pupae were calculated using the following formula: 4/3π ( L/2 ) ( W/2 ) 2 ( L , length; W , width ) . Brains were dissected from 3rd instar wandering larvae , fixed in 4% formaldehyde-PBS for 30 min at RT and washed in PBT buffer ( 0 . 3% Triton X 100 in PBS ) . Following blocking for 2 h at RT in 5% bovine serum albumin ( BSA ) /PBT , samples were incubated with respective primary antibodies at 4°C o/n . Samples were washed three times ( 15 min each ) with PBT and incubated with respective secondary antibodies for 2 h at RT . After three washes in PBT , brains were mounted in Vectashield Mounting Medium ( Vector laboratories ) . Fluorescence images were acquired using a Leica TCS SP5 MP SMD FLIM confocal laser scanning microscope . Antibodies used: rat anti-dILP2 [5] , rabbit anti-dILP2 [48] , guinea pig anti-dILP2 ( see below ) , rabbit anti-dILP5 ( 5 ) , anti-rat Alexa fluor 633 ( Invitrogen ) , anti-rabbit Alexa fluor 633 ( Invitrogen ) and anti-guinea pig Alexa fluor 633 ( Invitrogen ) . Confocal images of each IPC cluster were taken using the same scan and laser power settings . Total signal from each cluster was quantified using ImageJ software ( NIH ) . Representative images were cropped and processed identically for all the samples of every experiment . Two synthetic peptides ( PHKRAMPGADSDLDA ) and ( AEVRRRTRQRQGI ) corresponding to amino acid sequence of dILP2 from 49–63 and 102–114 amino acids respectively were produced by Storkbio Ltd . ( Estonia ) . These peptides were mixed and used as immunogens to raise polyclonal antibodies in Guinea pigs by Storkbio Ltd . RNA extraction from larval brains: Equal numbers of L1 larvae were collected and grown at 29°C . 10 brains from third instar non-wandering larvae ( approx . 85 hours AED ) were dissected and RNA was extracted using Nucleospin RNA XS kit ( Macherey-Nagel ) according to the manufacturer's protocol . RNA extraction from whole larvae: equal numbers of L1 larvae were collected and grown at 29°C for 72 h prior to harvesting ( 96 h AED ) . ≥5 third instar larvae per sample in minimum of three replicates were homogenized and RNA was extracted using Nucleospin RNA II kit ( Macherey-Nagel ) according to the manufacturer's protocol . RNA was reverse transcribed using RevertAid H Minus First Strand cDNA Synthesis Kit ( Thermo Scientific ) . qPCR was performed using Maxima SYBR Green qPCR Master Mix ( 2X ) ( Fermentas ) with Light cycler 480 Real-Time PCR System ( Roche ) . At least three biological replicates were used for each genotype and at least three technical replicates were used for each biological replicate . Primers for quantitative RT-PCR are presented as Table S3 . In situ hybridization was performed as described previously [49] . Briefly , antisense and sense probes were synthetized by T3/T7 RNA polymerase-dependent in vitro transcription reaction ( Promega ) in the presence of digoxigenin-UTP ( Roche ) . Mid-3rd instar larval tissues were fixed using 4% formaldehyde and they were hybridized with the labeled probes for 16 h at 55°C . The hybridization was visualized by using alkaline phosphatase-conjugated anti-digoxigenin Fab fragments ( 1∶3 , 000; Roche ) in the presence of NBT and BCIP substrates ( Promega ) . Primers for the erk7 probes were as follows: 5′-TAATACGACTCACTATAGGGCTCAAGAGCGACGCATTCAA-3′ 5′-ATTAACCCTCACTAAAGGGATACTGGTCCACGTCGTAGCG-3′
|
Ribosome biogenesis is a major consumer of cellular energy and a rate-limiting process during cell growth . The ribosome biogenesis pathway is tightly connected with signaling pathways that regulate tissue growth . For example , nutrient-regulated signaling cues adjust the rate of ribosome biogenesis . On the other hand , the process of ribosome biogenesis is closely monitored by so-called surveillance mechanisms . The best-known ribosome surveillance factor is the transcription factor and tumor suppressor p53 . In proliferating cells , activation of p53 upon disturbed ribosome biogenesis leads to cell cycle arrest and inhibition of proliferation . Here we show that ribosome surveillance not only regulates growth locally in proliferating cells , but is also coupled to hormonal growth control through regulation of insulin like peptide ( dILPs ) secretion . We observed that inhibition of ribosome biogenesis in the Drosophila insulin-producing cells generates a strong cell autonomous signal to inhibit dILP secretion . We identify two downstream effectors of this ribosome surveillance response by showing that p53 as well as an atypical MAP kinase ERK7 are mediators of the inhibition of dILP secretion . We also provide evidence that this ribosome surveillance mechanism contributes to nutrient-dependent regulation of dILP secretion .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"developmental",
"biology",
"animal",
"genetics",
"model",
"organisms",
"genetic",
"screens",
"gene",
"identification",
"and",
"analysis",
"genetics",
"biology",
"and",
"life",
"sciences",
"research",
"and",
"analysis",
"methods",
"gene",
"function"
] |
2014
|
p53- and ERK7-Dependent Ribosome Surveillance Response Regulates Drosophila Insulin-Like Peptide Secretion
|
Adenylyl cyclase ( AC ) is an important messenger involved in G-protein-coupled-receptor signal transduction pathways , which is a well-known target for drug development . AC is regulated by activated stimulatory ( Gαs ) and inhibitory ( Gαi ) G proteins in the cytosol . Although experimental studies have shown that these Gα subunits can stimulate or inhibit AC’s function in a non-competitive way , it is not well understood what the difference is in their mode of action as both Gα subunits appear structurally very similar in a non-lipidated state . However , a significant difference between Gαs and Gαi is that while Gαs does not require any lipidation in order to stimulate AC , N-terminal myristoylation is crucial for Gαi’s inhibitory function as AC is not inhibited by non-myristoylated Gαi . At present , only the conformation of the complex including Gαs and AC has been resolved via X-ray crystallography . Therefore , understanding the interaction between Gαi and AC is important as it will provide more insight into the unknown mechanism of AC regulation . This study demonstrates via classical molecular dynamics simulations that the myristoylated Gαi1 structure is able to interact with apo adenylyl cyclase type 5 in a way that causes inhibition of the catalytic function of the enzyme , suggesting that Gα lipidation could play a crucial role in AC regulation and in regulating G protein function by affecting Gαi’s active conformation .
Many proteins are involved in cell communication of which one type is the G-protein-coupled receptor ( GPCR ) , embedded in the membrane . GPCRs are part of a major signalling pathway , the GPCR signal transduction pathway , which enables the transfer of a signal from the extracellular region to the intracellular side and is a key target for drug development . A large diversity of GPCRs can be found in nature as about 800 human genes are involved in storing different types of GPCRs that can interact with neurotransmitters , hormones or exogenous ligands , for example [1] . In the cytosol , G proteins , composed of an α , β and γ subunit , are the first interaction partner of activated GPCRs . When a heterotrimeric G protein is activated by a GPCR , the trimer dissociates , resulting in an α subunit and a βγ dimer [2] . Activated Gα subunits transport the signal from the membrane to other regions of the cell by stimulating or inhibiting reactions via protein-protein interactions . Besides direct activation by GPCRs , the function of G proteins can also be influenced by other environmental factors , such as lipidation . Permanent N-myristoylation , for instance , is known to change the structure and function of the inhibitory G-protein subunit Gαi1 in its active GTP-bound state [3–6] . While a wide range of GPCRs exists , a relatively low diversity is present in the G protein family , e . g . in the human body . The human body includes only a relatively small variety of 21 α , 6 β and 12 γ subunits [1] . The Gα subunits are divided into four major subfamilies based on their sequence homology and function [7]: stimulatory Gαs , inhibitory Gαi , Gαq and Gα12 [8 , 9] . Overall the structures of the Gα subfamilies are similar ( S1 Fig , Fig 1 ) , including a Ras domain and an alpha helical ( AH ) domain . The Ras domain is present in all members of the G-protein superfamily and can perform hydrolysis of GTP to GDP during deactivation of the Gα subunit [10] . In addition , the domain includes an interaction site for GPCRs as well as regions that can interact with the βγ dimer . Moreover , the Ras domain can also undergo lipidation . Except for Gαt , all Gα proteins are able to reversibly bind a palmitate to their N-terminal helix . Besides palmitoylation , Gαi can also permanently bind a myristoyl moiety to the N-terminus that appears to be crucial for the function of the subunit ( Fig 1C ) [4 , 5 , 9 , 11] . The AH domain is unique to the Gα subfamilies , which is composed of six α helices and interacts with the Ras domain when GTP or GDP is present ( Fig 1C ) . However , this interaction between the AH and Ras domain is weakened when a nucleotide is absent in Gα’s active site [12–14] . The high structural similarity among members of the Gα subfamilies is illustrated by aligning the X-ray structures of stimulatory Gαs and inhibitory Gαi1 , resulting in a root mean square deviation ( RMSD ) of only 1 . 07 Å between the Cα atoms of the two structures ( Fig 1A ) [15 , 16] . Hence , from a comparison of the structures it is difficult to conclude what the origin is of their inverse action , i . e . , how the structure can be related to a stimulatory respectively an inhibitory effect . An example of a protein in which both Gαi and Gαs are important for regulation is adenylyl cyclase ( AC ) . Ten isoforms of AC are known of which nine are membrane-bound ( AC1-9 ) and one is soluble ( sAC ) [17] . These different types of AC are found throughout the body in different concentrations . AC5 , for instance , is present in high quantities in the brain , the spinal cord and the heart , and is associated to congestive heart failure and pain perception [18 , 19] . G proteins have the ability to either stimulate ( Gs ) or inhibit ( Gi ) adenylyl cyclases’ conversion of adenosine triphosphate ( ATP ) to cyclic adenosine monophosphate ( cAMP ) and pyrophosphate [20 , 21] . ACs consist of two membrane-bound regions , each built from six trans-membrane domains , and a catalytic region in the cytosol that includes two pseudo-symmetric domains , C1 and C2 ( Fig 1 ) [22] . GTP-bound Gαs is known to bind to the C2 domain for which the interaction site is known from X-ray structures of Gαs interacting with AC ( Fig 1B ) [23] . Such data is absent for the case of Gαi . In the absence of direct experimental information , a putative interaction site of GTP-bound Gαi has been suggested in analogy to the known structure of the complex of Gαs and AC ( Gαs:AC ) as the pseudo-symmetric site on the C1 domain ( Fig 1B ) . However , how the interaction of Gαi on the C1 domain should induce inhibition is not obvious [5] . Furthermore , since with this hypothesis the interaction sites of Gαi1 and Gαs are highly similar in addition to their structures , it is unclear how the α subunits can differentiate the two binding sites on AC and what the cause is of the stimulatory versus the inhibitory effect induced by the subunits . A factor that could play an important role in differentiating the action of Gαi and Gαs is the difference in lipidation of both subunits . Although the X-ray structures in the Protein Data Base ( PDB ) [24] of the active inhibitory and stimulatory Gα subunits tightly align , the N-terminus , which is not resolved for Gαi or Gαs , is not myristoylated during the expression process of Gαi as lipidation can hinder crystallisation [4] . Hence , it is not clear to what extent the missing N-terminal myristoyl moiety affects the Gαi structure of the remaining protein while the bound myristoyl group has been known to be crucial for Gαi’s conformation and function as the ability to interact with AC5 is abolished upon removal of the myristate [4–6] . Classical molecular dynamics ( MD ) simulations of myristoylated GTP-bound Gαi1 , Gαi1myr , demonstrate the stability of the myristoyl moiety on the Ras domain due to a hydrophobic pocket formed by β2-β3 , α1 and the C-terminus α5 ( Fig 1C ) and show that myristoylation can have a significant effect on the conformation of the subunit [25] . The findings suggest the possibility of an alternative novel interaction mode and open up new possibilities for selective interactions with AC . This is because the found structural changes in the classical MD simulations of Gαi1myr [25] suggest that the subunit will not be able to interact with C1 as Gαs interacts with C2 . Here , we investigate the interaction between Gαi1myr and AC , using classical MD simulations . To this end , the initial structure of Gαi1myr was taken from reference [25] in which a 2 μs classical MD simulation of Gαi1myr is described . Gαi1myr can inhibit only particular isoforms of AC: AC1 , AC5 , AC6 [26] . In this study AC5 is used because X-ray structures of AC’s catalytic domains are composed of isoforms AC2 and AC5 . Ca . 16 AC structures can be found in the PDB with different resolutions and/or crystallisation conditions . All available structures have been co-crystallised with a Gαs subunit and correspond therefore to stimulated conformations at various levels of activation , depending on the nature of bound cofactors ( e . g . cofactor-free complex of AC , substrate-bound AC complex ) . When AC5 becomes active , roughly three conformational options are possible: a complex of Gαs and AC5 , Gαs:AC5 , Gαs in complex with ATP-bound AC5 , Gαs:AC5 ( ATP ) , or a complex of Gαs and AC5 bound to the reaction products cAMP and pyrophosphate , Gαs:AC5 ( cAMP ) . Currently , it is not known which one of these forms is most likely to interact with Gαi1myr , or if Gαi1myr can inhibit all of them . In this study , the structure of the AC5 protein was taken from a crystal structure of the cofactor-free Gαs:AC5 complex . This apo AC5 structure was used as it could provide insight into Gαi1myr’s inhibitory effect on a stimulated conformation of AC5 in the absence of ATP . The selected AC5 structure was employed to build a Gαi1myr:AC5 complex ( Fig 1C ) and to explore if the binding of Gαi1myr is able to affect the active conformation initially induced by Gαs . The absence of ATP in the active site provides the opportunity to investigate Gαi1myr’s ability to prevent the formation of AC’s fully activated form by altering AC’s conformation unfavourably prior to substrate association . In order to verify which changes are due to the interaction of AC5 with Gαi1myr and which alterations are a result of the removal of Gαs , a second simulation of AC5 , with the Gαs subunit removed , was performed on the same time scale as the Gαi1myr:AC5 complex . Hence , in this study the impact of the presence of myristoylated Gαi on the function of AC5 is explored via investigating the conformational features of the Gαi1myr:AC5 and the free AC5 complex ( a system that only includes AC’s catalytic region in solution ) in comparison with the Gαs:AC X-ray structure . The Gαi1myr:AC5 complex has been obtained via docking the Gαi1myr structure on to the C1 domain of AC5 . Already the initial docking results confirm the possible importance of the myristoyl-induced structural changes of Gαi1myr as a new interaction mode for Gαi1myr could be identified . The comparison of the performed classical MD simulation ( 2 . 5 μs ) of the Gαi1myr:AC5 complex and the free AC5 system suggest two possible ways of AC inhibition in its apo form . First , Gαi1myr seems to inhibit AC’s conversion of ATP to cAMP by preventing active-site formation as the Gαi1myr subunit perturbs the conformation of the active site at the C1/C2 interface . Second , the effect of Gαi1myr on the AC structure leads to a closed conformation of the Gαs binding site on C2 , decreasing the probability of Gαs association and thus of a counter-balancing re-stimulation of the AC5 activity . Taken together , the observed events lead to a suggestion for a putative Gαi1myr inhibition mechanism of apo AC5 in which lipidation is crucial for Gαi1myr’s function and its protein-protein interactions . Hence , the results of this study provide a possible indication that lipidation could play a significant role in regulating G protein function and therefore could impact signal transduction in G protein mediated pathways [4–6] .
The PDB structure 1AZS , including the Gαs:AC complex with AC in the apo form , was used as a template , including 1AZS’s C1 and C2 domain , for the initial AC5 structure of Rattus norvegicus ( UniprotKB Q04400 ) [27–29] . The structure of the Rattus norvegicus Gαi1myr subunit ( UniprotKB P10824 ) interacting with GTP and Mg2+ was taken from reference [25] ( S2 and S3 Figs ) . The HADDOCK web server [30] was used for docking ten conformations of Gαi1myr on the catalytic domains of AC5 of Rattus norvegicus . The Gαi1myr snapshots were extracted at the end of the Gαi1myr classical MD trajectory ( around 1 . 9 μs ) discussed in reference [25] , with a time interval of 0 . 5 ns . The active region of Gαi1 was defined in HADDOCK as a large part of the AH domain ( 112-167 ) , the switch I region ( 175-189 ) and the switch II region ( 200-220 ) , allowing for a large unbiased area on the Gαi1myr protein surface to be taken into account during docking . The active region of AC5’s C1 domain was defined as the α1 helix ( 479-490 ) and the C-terminal region of the α3 helix ( 554-561 ) because experimentally it has been found that Gαi1myr is unable to interact with C2 and its main interactions with AC are with the C1 domain [5] . Passive residues , residues that could take part in protein-protein interaction , were defined as residues around the active residues that are on the protein surface and within a radius of 6 . 5 Å of any active residue [30] . The initial Gαi1myr:AC5 complex for the classical MD simulations was selected based on ( 1 ) the absence of overlap between the C2 domain and Gαi1myr , ( 2 ) no overlap with Gαi1myr’s GTP binding region and the interaction site of Gαi1myr with C1 and ( 3 ) presence of similar complexes in the top-ten docking results of the docking calculations performed for all ten Gαi1myr snapshots . The first property of the selection criteria is important since Gαi1myr is unable to interact with C2 [5] . The second criterium has been defined since GTP is located in the active site of Gαi1myr in the classical MD simulations , but was not incorporated in the docking procedure because this is not possible in HADDOCK . Therefore , no overlap between the GTP binding site and the C1 domain should be present in the docking result as otherwise the GTP molecule will not be able to fit in Gαi1myr’s active site . The last criterium is the presence of similar Gαi1myr:AC5 complexes of the selected complex in all top-ten docking results which increases the probability that complexation of the two proteins is not conformation specific , but is robust as similar complexes can be obtained using different conformations of Gαi1myr . The Gαi1myr:AC5 complex was used to simulate the protein complex for 2 . 5 μs at 310 K and 1 bar using a Nosé-Hoover thermostat and an isotropic Parrinello-Rahman barostat . In the active site of Gαi1myr one Mg2+ ion and a GTP molecule are present . In order to closer mimic an AC5 system with which ATP or a product such as pyrophosphate would be able to interact , a Mg2+ ion was added to the active site of AC5 ( see S1 Appendix ) . Additionally , about 68 000 water molecules and 150 mM KCl are present in the simulated system . The force fields used for the protein and the water molecules are AMBER99SB [31] and TIP3P [32] , which were employed by Gromacs 4 . 6 . 6 [33 , 34] to perform the runs . For GTP , the force field generated by Meagher et al . was used [35] . The adjusted force field parameters for the K+ ions and the Cl- ions were taken from Joung et al . [36] . The Mg2+ ion parameters originated from Allnér et al . [37] and the parameter set for the myristoyl group was taken from reference [25] . The charges for the myristoyl group were obtained with Gaussian 09 [38] based on Hartree Fock calculations in combination with a 6-31G* basis set and using the AMBER RESP procedure [39] . Appropriate atom types from the AMBER99SB force field were selected to complete the myristoyl description . Electrostatic interactions were calculated with the Ewald particle mesh method with a real space cutoff of 12 Å . Bonds involving hydrogen atoms were constrained using the LINCS algorithm [40] . The time integration step was set to 2 fs . The free AC5 system was simulated with the same setup as the Gαi1myr:AC5 complex . The system was solvated in 30 000 water molecules and a 150 mM KCl concentration . The initial location of the Mg2+ ion in the active site of the enzyme was the same as in the Gαi1myr:AC5 complex system . Multiprot [41] and VMD [42] were used to align protein structures . Uniprot [43] was used to align protein sequences . Images were prepared with VMD [42] .
The initial conformation of the Gαi1myr:AC5 complex suggests that Gαi1myr’s proposed interaction site ( see S1 Appendix ) affects the conformation of C1 in a different way than Gαs stabilises the C2 domain ( Figs 1 , 2 and S3 Fig ) . Unlike Gαs , Gαi1myr is not located between the helices of AC5’s catalytic domain , but appears to clamp the C1 domain into its inactive conformation . Gαi1myr is positioned around AC5’s α3 , interacting with α1 , α2 , and α3 via its switch I , II and III region together with the C-terminal domain of αB ( Fig 2 and S3 Fig ) . Since C1’s α1 helix appears to decrease its distance with respect to the C2 domain when an ATP analog , adenosine 5- ( α-thio ) -triphosphate ( ATPαS ) , is present in the active site ( S2 Fig ) , the interactions between Gαi1myr and C1’s α1 in the Gαi1myr:AC5 complex could suggest that one way by which Gαi1myr is able to inhibit ATP’s conversion is by preventing C1’s α1 to rearrange upon ATP binding . The simulation of Gαi1myr:AC5 in comparison with the free AC5 trajectory and the Gαs:AC X-ray structure demonstrate that the first step in decreasing AC5’s activity in the apo form is the relocation of the β7-β8 loop ( Fig 7 , step one ) . In fact , the β7-β8 loop seems to have an important role for the stimulatory response since the presence of Gαs leads to the stabilisation of the loop , forming ATP’s binding site ( Fig 7 , starting conformation of AC in left panel ) [23] . This loop conformation is lost as soon as Gαs is absent , as observed for both free AC5 and Gαi1myr:AC5 . In step two of Fig 7 the Gαi1myr:AC5 complex undergoes a rearrangement in the C2 domain ( absent in free AC5 ) , which leads to a further perturbation of AC5’s active site . The classical molecular dynamics simulations also show that in the presence of Gαi1myr , there appears to be a decrease in probability for Gαs association ( Fig 7 , step 3 and Fig 6 ) . Hence , through these rearrangements Gαi1myr could deactivate apo AC5 as well as decrease the probability of reactivation via Gαs . The results of this study suggest that Gαi1myr deactivates the apo form of adenylyl cyclase type 5 via constraining C1’s active site region . Inhibition and stimulation of AC5 appear to follow different pathways . While Gαs binds between the helices of C2 , increasing the stability of the C1:C2 dimer , Gαi1myr is able to clamp the helices of the C1 domain , promoting an inactive conformation of AC5’s catalytic domains and a possible decrease in affinity for Gαs on the C2 domain . Structurally , Gαs and non-myristoylated Gαi1 are very similar , however , when myristoylation has taken place on the N-terminus of Gαi1 , the conformation of the subunit changes drastically , leading to a structure that differentiates itself from the active Gαs subunit and enables the protein to function in an inhibitory fashion as is shown via the presented classical MD simulations . Hence , in line with experimental studies , myristoylation appears to be crucial for Gi’s function and demonstrates how important even relatively small changes to a protein structure can be for its function .
|
Communication between cells is essential for the survival of any multicellular organism . When these mechanisms cannot function properly , diseases can occur such as heart failure or Parkinson’s disease . Understanding cell communication is therefore crucial for drug development . Important proteins in cellular signalling are the ones that initiate mechanisms in the cell after the signal of an extracellular trigger is transported from outside to inside the cell . G proteins ( GPs ) are an example of such proteins . Experimental studies have shown that GPs can perform stimulatory or inhibitory functions , however , it is not well understood what the difference is in their mode of action , especially as they are structurally very similar . Adenylyl cyclase ( AC ) is an enzyme which can be stimulated or inhibited by GPs , depending on which type of GP is active . Hence , AC is a good candidate for investigating the difference in function between GPs . However , only the structure of the stimulatory GP interacting with AC is known . Here , we investigate for the first time the effect of the interaction of an inhibitory GP with AC via classical molecular dynamics simulations in order to obtain a better understanding of the difference between stimulatory and inhibitory GP association and AC regulation .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"and",
"discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"molecular",
"dynamics",
"diagnostic",
"radiology",
"crystal",
"structure",
"enzymes",
"condensed",
"matter",
"physics",
"enzymology",
"protein",
"structure",
"lyases",
"crystallography",
"g",
"protein",
"coupled",
"receptors",
"research",
"and",
"analysis",
"methods",
"solid",
"state",
"physics",
"imaging",
"techniques",
"proteins",
"chemistry",
"transmembrane",
"receptors",
"molecular",
"biology",
"physics",
"protein",
"structure",
"comparison",
"biochemistry",
"signal",
"transduction",
"biochemical",
"simulations",
"radiology",
"and",
"imaging",
"diagnostic",
"medicine",
"cell",
"biology",
"adenylyl",
"cyclase",
"genitourinary",
"imaging",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"computational",
"chemistry",
"computational",
"biology",
"macromolecular",
"structure",
"analysis"
] |
2017
|
Exploring the inhibition mechanism of adenylyl cyclase type 5 by n-terminal myristoylated Gαi1
|
Single-cell RNA sequencing ( scRNA-seq ) has become a powerful tool for the systematic investigation of cellular diversity . As a number of computational tools have been developed to identify and visualize cell populations within a single scRNA-seq dataset , there is a need for methods to quantitatively and statistically define proportional shifts in cell population structures across datasets , such as expansion or shrinkage or emergence or disappearance of cell populations . Here we present sc-UniFrac , a framework to statistically quantify compositional diversity in cell populations between single-cell transcriptome landscapes . sc-UniFrac enables sensitive and robust quantification in simulated and experimental datasets in terms of both population identity and quantity . We have demonstrated the utility of sc-UniFrac in multiple applications , including assessment of biological and technical replicates , classification of tissue phenotypes and regional specification , identification and definition of altered cell infiltrates in tumorigenesis , and benchmarking batch-correction tools . sc-UniFrac provides a framework for quantifying diversity or alterations in cell populations across conditions and has broad utility for gaining insight into tissue-level perturbations at the single-cell resolution .
Single-cell sequencing technologies enable profiling of hundreds to thousands of individual cells from a tissue composed of diverse cell types [1–8] . The rapid advances in single-cell technologies have led to a proliferation of novel computational tools . Current analyses mainly focus on identification of cell populations and transitional trajectories from data obtained from a single sample ( reviewed in [9–12] ) . While these tools can shed light on complex biological processes within a given sample [13] , it is becoming apparent that the power of single-cell technologies lies in multi-sample experiments , for which the response of single cells—in terms of identity and quantity following multiple perturbations—can be assessed . This poses a new statistical challenge: given that each sample is composed of transcriptomes from hundreds to thousands of individual cells , how does one compare across different samples to statistically assess population diversity and to detect population changes in an unbiased manner ? As an increasing number of studies are generating single-cell RNA sequencing ( scRNA-seq ) data from multiple samples , there is an unmet need for a statistical framework that enables quantitative comparisons across different single-cell landscapes . Such a framework would have wide application from quantitatively uncovering cell population changes and assessing batch effect correction methods to classifying disease subtypes based upon single-cell landscapes . There currently exists a paucity of approaches to determine cellular composition similarities and differences between samples . Citrus is a supervised approach for identifying cell populations that are significantly different between specified outcomes , and its goal is distinct from unsupervised comparison of samples based on similarities [14] . Another approach uses the Wasserstein metric known as Earth Mover’s Distance ( EMD ) , which is a measure of the distance between two probability distributions over a certain data space [15] . Briefly , EMD partitions the entire data space into bins and measures the cost of transfer of data points from one distribution across these bins to resemble the other distribution . Orlova and colleagues applied EMD across datasets to quantify the similarity of two cell populations by measuring the distance between expression distributions in two-dimensional marker space [16] . However , partition of data spaces into bins has an exponential computational cost as the number of dimensions increases , limiting EMD use to one- or two-dimensional data spaces . Although there are multiple dimension-reduction approaches for high-dimensional data [11 , 17 , 18] , analysis in a customizable , unrestricted number of dimensions would be preferred , especially for scRNA-seq datasets with thousands of native dimensions . We have developed the p-Creode score , which determines the similarities between p-Creode trajectories derived from a multidimensional single-cell landscape [19] . Most recently , cellAlign was developed to align single-cell trajectories using dynamic time warping , and an alignment-based distance was defined to evaluate similarities between trajectories [20] . However , these approaches are limited to datasets in which trajectories can be derived from continuous single-cell data and are not generalizable to all data distributions . Currently , the most common strategy for assessing similarities between single-cell landscapes remains a visual evaluation of the degree of “mixing” of data points when two or more samples are analyzed together on a t-distributed stochastic neighbor embedding ( t-SNE ) plot [2] . Two methods have recently been developed to use k-nearest neighbors to characterize this degree of intermixing [21 , 22] . Both methods use the simple assumption that k-nearest neighbors of each cell should have the same distribution of sample labels as the full dataset if the datasets are well mixed . We were originally inspired by some of the early single-cell work in which similarities between replicates can be qualitatively evaluated by the degree of mixing of hierarchical clusters between replicates [7] . Thus , by deriving a quantitative measure to compare between hierarchical trees generated by clustering , we can obtain a corresponding quantitative , statistically testable metric to compare cell population diversity between single-cell landscapes . Multiple metrics to measure similarity between trees have been proposed , such as Baker’s gamma index [23] , which we used previously to determine the similarity between signaling modules [24] . UniFrac is a distance metric originally devised to compute differences between microbial communities by incorporating phylogenetic information [25] . UniFrac provides a qualitative measure , which is calculated as the fraction of the total unshared branch lengths . A weighted version of UniFrac is a quantitative measure , such that branch lengths are weighted by the relative abundance of each taxon [26] . We created a workflow , called sc-UniFrac , that enables the application of the weighted UniFrac statistical framework on single-cell data to identify and characterize cellular diversity that distinguishes single-cell landscapes . Pairwise comparisons using sc-UniFrac can easily be extended to multi-sample experimental designs that are increasingly common in single-cell studies . sc-UniFrac compares diversity based on transcriptome similarities of single cells and is more powerful than intermixing methods based on its accounting of both the global and local structures of the data . We demonstrated the utility of sc-UniFrac in quantifying similarities between simulated and real sc-RNAseq datasets , for which the ground truth of similarities between samples is known . We also successfully applied sc-UniFrac to detect technical effects from replicate samples , assess the performance of batch-correcting methods , and implicate the sources of technical variation , although sc-UniFrac itself does not correct for these effects . We envision that quantitative metrics such as sc-UniFrac will find increasing utility as the field continues to generate a greater number of sc-RNAseq datasets from different conditions . sc-UniFrac will greatly facilitate single-cell studies , including those aimed at deciphering how cell populations respond to perturbations or tracking the evolution of cell populations during disease progression .
For quantifying cell diversity differences in single-cell landscapes , we borrowed the UniFrac concept from the microbiome field . UniFrac is a distance metric for quantifying differences in phylogenetic diversity between ecological landscapes . Instead of operating on phylogenetic trees that describe microbial diversity between two samples , sc-UniFrac has been adopted for single-cell research and builds hierarchical trees from clustering analyte profiles ( e . g . , transcriptome ) of single cells combined from two datasets . The purpose of clustering was not to separate cells into distinct groups , as performed traditionally for single-cell transcriptome datasets; instead , the hierarchical tree in sc-UniFrac is only to discern potential structures within the data . Thus , clustering can be performed with any method and can be applied to data with any distributions . The hierarchical tree can be constructed post hoc using distances between cluster centroids . The clustering tree , encompassing the structural features of cell subpopulations , is used for calculating the weighted UniFrac distance . sc-UniFrac is calculated by weighting relative abundance of samples assigned to each branch , as well as the branch length that denotes the distance between cluster centroids ( Fig 1A ) . A permutation test , conducted by randomizing the sample labels of the cells without changing the tree topology , is used to calculate statistical significance of the overall sc-UniFrac distance ( i . e . , to test whether the cell populational structures between the two samples are identical ) ( see Methods ) . For single-cell data , it is very important to identify cell populations altered across conditions , which are derived from branches that have significant proportion shifts between two samples . Instead of relying on exact , nonzero proportional differences , which would find every branch and easily be skewed by outliers , sc-UniFrac leveraged the above permutation test to identify branches whose proportional shifts cannot occur by chance alone . This procedure corrects for the algorithm’s sensitivity to noisy outliers prevalent in scRNA-seq experiments . After identifying cell populations that drive the compositional difference between single-cell transcriptome landscapes , sc-UniFrac detects gene signatures that mark these cell populations . Finally , sc-UniFrac predicts the potential identities of these populations by matching individual cell signatures to cell types from reference atlases . sc-UniFrac can operate in two modes: ( 1 ) pairwise comparisons and ( 2 ) extension of the pairwise approach to a multi-sample experimental design . The general workflow of the sc-UniFrac pipeline is shown in Fig 1B . Note that sc-UniFrac operates agnostically to technical or biological effects in the datasets and can be used to evaluate either depending on the input data . To evaluate technical effects , control replicate samples over multiple batches would be used as the input . To evaluate biological differences , it is assumed that technical variation in the input datasets has been well controlled for or batch correction has already been applied in previous processing steps . To evaluate the performance of sc-UniFrac in quantifying the compositional difference between single-cell landscapes , we first applied the method to experimental datasets in which population structure can be precisely controlled . Two 1 , 000-cell populations were generated by sampling from the CD8 and CD4 T-cell populations , respectively , from the T-cell development dataset of the mouse thymus ( S1 Data ) [25] . This process was repeated 50 times to evaluate the robustness of sc-UniFrac . Comparing the CD8 versus CD4 populations ( 1 , 000-cell sample from each population ) using sc-UniFrac resulted in a large sc-UniFrac distance , indicating , as expected , completely different cell populations ( Fig 2A—red arrow , S1A Fig ) . In contrast , comparing two 1 , 000-cell populations resampled from CD8 cells revealed that they possess the same population structures with a median distance of 0 ( Fig 2A—green arrow , S1B Fig ) . We then evaluated the performance of sc-UniFrac in a simulation experiment in which we constructed a series of paired samples with a gradation of proportional shift in cell populations . For each pair , one sample included only CD8 cells ( N1 ) , while the other was composed of proportional mixtures of CD4 and CD8 cells ( N2 ) , starting from 0% of CD4 cells ( no shift ) to 100% of CD4 cells ( complete shift ) . The distance of sc-UniFrac progressively increased as proportional shifts became larger ( Fig 2A and S1C and S1D Fig ) . Among the 50 resampled runs , less than 0 . 2% of the sc-UniFrac distances generated were significantly different ( identified as false positives ) when two samples were identical ( 0% proportional shift ) , while over 95% of the distances were significantly different even when there was as little as 2% of CD4 cells mixed in with the CD8 cells ( Fig 2B ) . These results signify that sc-UniFrac is sensitive and specific for detecting minute shifts in population structure . Next , using this controlled sampling scheme , we evaluated the impact of changing various parameters on the performance of sc-UniFrac . First , we altered the k parameter for dividing the data into k subpopulations for analysis , which tunes the resolution by which sc-UniFrac analyzes the datasets . The quantitative ability and sensitivity of sc-UniFrac were robust as long as k was not exceedingly low ( k > 3 ) ( Fig 2C and 2D ) . At k ≤ 3 , single-cell datasets are represented by three major groups or less , reducing the resolution of sc-UniFrac for detecting differences in single-cell level diversity . As k increases , the performance of sc-UniFrac continues to improve slowly until reaching an asymptote at the expense of higher computational costs . When k is extremely high , sc-UniFrac might be distracted by unimportant details ( noise ) and fail to capture global structures . Thus , we conclude that the performance of sc-UniFrac is minimally affected by the selection of k ( as long as it is not too low or extremely high ) . We recommend a k in the range of 10 to 30 depending on the heterogeneity of the data . Second , we assessed the effect of imbalanced dataset sizes . Keeping the size of sample 1 ( N1 ) constant at 1 , 000 cells , we altered the size of sample 2 ( N2 ) during resampling while maintaining the same population structure . The sc-UniFrac method was observed to be robust with respect to dataset size imbalances , with only minor losses in sensitivity at larger imbalances ( detection limit of 5% instead of 2% ) ( Fig 2E and 2F and S2 Fig ) . These results demonstrate the robustness of sc-UniFrac , whose performance is independent of input parameters . The same analysis using scRNA-seq data demonstrated similar results ( S3 Fig ) . Here , we extracted erythrocyte and myeloid cell progenitor cell populations from the Paul and colleagues dataset ( S2 and S3 Data ) [26] . Using the erythrocyte as a base population ( N = 500 cells ) , we progressively mixed in the progenitor cells to construct simulated datasets in a manner similar to the CD4/CD8 analysis . The general conclusions are the same , with two minor differences . First , this dataset is more robust to parameter changes due to the increased distinctiveness between the two cell populations . We speculate that better separation can be achieved either through true biological difference or by having more dimensions to define cell populations in scRNA-seq data compared with Cytometry Time-of-Flight ( CyTOF ) data . Second , sensitivity to detect dissimilarity is slightly decreased ( from 2% to 5% disparate cells for k = 10 , e . g . ) due to smaller dataset sizes . Nevertheless , the robustness of the algorithm to the k parameter and dataset size imbalance still holds in this analysis . To demonstrate the utility of sc-UniFrac on scRNA-seq data , we generated a series of scRNA-seq datasets with known similarities and differences using inDrop ( 1Cell Bio; Cambridge , MA ) sequencing [1] . These datasets included tissue samples from mouse colonic epithelium , consisting of both technical and biological replicates , as well as biological replicate samples of mouse embryonic pancreatic islets collected at embryonic day 14 . 5 ( E14 . 5 ) . Three technical replicates ( colon1_1 , colon1_2 , colon1_3 ) consisted of multiple single-cell fractions collected from 1 mouse , whose libraries were constructed and sequenced on separate days . Three biological replicates ( colon1 , colon2 , colon3 ) consisted of samples collected from different mice on different days but underwent library preparation and sequencing together . One sample ( colon1 ) has both technical and biological replicates . Two traditional strategies to assess the reproducibility of scRNA-seq datasets—for both biological and technical replicates—were performed . One strategy is to compare the median levels of all genes expressed between every sample pair using Spearman correlation analysis . The correlation analysis demonstrated that the technical replicates were more similar among each other ( mean R = 0 . 89 ± SD 0 . 04 ) compared with samples among biological replicates ( mean R = 0 . 80 ± SD 0 . 07 ) ( S4 Fig ) . As expected , the embryonic islets displayed a median gene expression that was the most different when compared to the adult colon ( mean R = 0 . 504 ± SD 0 . 07 ) . These results are consistent with the expected similarity among different conditions with technical replicates being more similar than biological replicates , which are more similar than the outgroup organ . While correlation analysis can quantify the degree of similarity in terms of the average transcriptional profile , it provides a very rough estimate and tends to be easily biased towards the dominant population in single-cell landscapes . The other strategy is to do a visual evaluation on how cells from multiple samples are intermixed on a t-SNE plot . Structural differences among different samples will be reflected in segregation of data points into separate clusters by sample , while data points from similar samples will appear together as mixed clusters . Visualization on a t-SNE plot showed the same result as median correlation analysis . Technical replicates appeared more intermixed within t-SNE clusters compared with biological replicates ( Fig 3A and 3B and S5A and S5B Fig ) . In contrast , pancreatic islet biological replicates segregate away from samples generated from both the technical and biological replicates of the colon ( Fig 3C ) . While t-SNE analysis describes the subpopulation structure of the samples , it is not quantitative in that similarities and differences were assessed subjectively by visualization . Compared with these two traditional strategies , sc-UniFrac provides an objective , precise , and unbiased metric to quantify compositional dissimilarities across scRNA-seq datasets while taking population structures into account . The calculated sc-UniFrac distance between the colonic and pancreatic islet datasets was 1 , the maximal obtainable distance demonstrating that the samples did not share any cell populations ( Fig 3D and S5C Fig ) . Much smaller , but significant , distances were observed among biological replicates of colonic datasets ( Fig 3D and S5C Fig , distance = 0 . 24–0 . 37 ) , suggesting that they share cell populations but proportional compositional difference can still be detected . Technical replicates appeared the most similar with sc-UniFrac being marginally small without statistical significance ( Fig 3D and S5C Fig , distance = 0 . 05–0 . 09 ) , suggesting that they are composed of almost identical data points . The ordering by similarity across samples was robust to the k parameter ( Fig 3E and 3F ) . A metric was defined to evaluate the power of sc-UniFrac for discriminating biological replicates from technical replicates ( discriminative ability ) by subtracting the smallest distance between biological replicates by the largest distance between technical replicates . A positive discriminative ability suggests that sc-UniFrac can discriminate technical from biological replicates . Notably , the k parameter again did not affect the discriminative ability except when k is very small ( k ≤ 2 ) ( Fig 3E and 3F ) . These results demonstrate the ability of sc-UniFrac to objectively and quantitatively determine dissimilarities between single-cell datasets , as seen by the ordering of samples by their expected similarities ( technical replicate > biological replicate > outgroup organ ) . We performed a comparison of sc-UniFrac to published methods for assessing single-cell landscape similarity , using the above dataset with known similarity ordering . First , we evaluated cellAlign , an algorithm for aligning two unbranched pseudotemporal trajectories using dynamic time warping [20] . The alignment-based distance was defined to evaluate the similarity of two trajectories . cellAlign requires the input to be continuous scRNA-seq data comprising single unbranched trajectories . To generate such an input , we manually selected cells that form unbranched data continua from stem cells to colonocytes for the colonic datasets as well as from endocrine progenitors to beta cells for the pancreatic islet datasets ( S4 Data ) . The cellAlign distance , similar to sc-UniFrac , revealed that colonic datasets can be distinguished and clustered away from pancreatic islet datasets , while biological and technical replicates of the colon cannot be clearly delineated ( Fig 4A ) . An example of a nonideal alignment between technical replicates compared with an alignment between biological replicates is shown ( Fig 4B and 4C ) . These results suggest that sc-UniFrac is more powerful for distinguishing differences between single-cell landscapes than cellAlign . The low sensitivity of cellAlign might be due to various parameters , such as dataset size imbalance or uneven sampling of datapoints along a trajectory , which we have not thoroughly tested here . Furthermore , the application of cellAlign is restricted to the very specific case of continuous data that form a single unbranched trajectory , thus limiting its generalizability . To extend orthogonal methodologic comparisons to sc-UniFrac , we also used p-Creode trajectory analysis , an algorithm by which a single-cell landscape composed of continuous cell-state data is represented as acyclic graphs to model transition trajectories [19] . The p-Creode score , developed to determine the topological similarity between graphs with differing nodes and edges , can be used to quantify dissimilarities of the trajectory graph outputs generated from different datasets . We revised the p-Creode scoring method to accommodate comparisons of graphs of difference sizes by interpolating between edges connecting nodes instead of directly matching node positions between the two test datasets ( S6A–S6C Fig and Methods ) . p-Creode was applied to each dataset for 100 times to generate consensus trajectories using data resampling . The modified p-Creode score was used as a distance metric for clustering cell-transitional trajectories created from the resampled datasets . Consistent with sc-UniFrac , p-Creode trajectories among technical replicates clustered together using the p-Creode score as a dissimilarity metric , while data from biological replicates were more disparate ( Fig 4D and 4E ) . As expected , organ specificity drove clustering when pancreatic islet data were added to the analysis , with all trajectories generated from colonic data clustering together away from pancreatic trajectories ( S6D and S7 Figs ) . Similar to cellAlign , p-Creode was designed for data that are distributed as a continuum , and not as distinct clusters . However , the p-Creode score can also be used to evaluate complex multi-branching trajectories in addition to linear ones . Nevertheless , comparisons can only be made for tissue systems that are transitioning , which is the case for both the adult colonic epithelium and embryonic pancreatic islets compared here . sc-UniFrac does not have this limitation because it can compare between datasets of any distribution , including continuous data , as well as discrete populations that are composed of cells from different lineages . Thus , sc-UniFrac has greater general utility than the above two trajectory comparison methods for determining dissimilarities of single-cell datasets in an unsupervised way without prior knowledge of the distribution of the data . While sc-UniFrac can statistically measure the population diversity between two single-cell landscapes , it also provides an easy and intuitive way to identify the cells that drive the differences . The distance-driving cells are either expanding or contracting populations or even newly emerging populations across conditions . Instead of relying on nonzero proportional differences between two samples , which would identify every branch and easily be skewed by outliers , sc-UniFrac uses the incorporated permutation significance tests to detect branches whose proportion shifts in cell populations cannot happen by chance alone ( Methods ) . The branches with significant proportional shifts between two samples are distance-driving cells . Illustrating this concept , sc-UniFrac was performed to demonstrate pairwise comparison of scRNA-seq datasets of colonic and pancreatic tissue with k = 10 . As expected , technical replicates of the colon with the smallest sc-UniFrac value have mostly shared branches between them , with only one branch with subtle proportional shifts ( Fig 5A ) . In contrast , comparison of the pancreatic and colonic datasets revealed no shared branches , with every unshared branch being highly significant ( Fig 5B ) . Evaluation of unshared branches can easily pinpoint cell groupings that contribute to sc-UniFrac . Here , we focus on group 10 , which was composed entirely of cells from the pancreatic sample . Supervised analysis of differential gene expression revealed the unique gene signatures of these cells compared to colonic populations , which can be identified by canonical marker genes ( e . g . , group 1 represents deep crypt secretory cells; 2 and 3 are colonocytes; 4 and 5 are goblet cells; 6 are intraepithelial lymphocytes ) ( S8A Fig ) . Projection of cells from group 10 onto reference cell type gene expression signatures from the Mouse Cell Atlas [3] revealed that individual cells mapped onto pancreatic acinar cells , duct cells , endocrine cells , and immune cells ( Fig 5C ) . These results demonstrate the utility of the branching feature of sc-UniFrac to statistically determine cell populations that drive differences between single-cell landscapes . Notably , none of the methods that we used above for comparison with sc-UniFrac can perform this task . While sc-UniFrac was able to derive the correct ordering of similarity between normal tissues , we next applied sc-UniFrac to more challenging cases to determine whether it can decipher meaningful dissimilarities in cell diversity arising from the same tissue . Cell type diversity is altered during the process of tumorigenesis as mutations alter signaling pathways to convert cells to abnormal states while , at the same time , additional cell types are recruited to the tumor microenvironment . Nevertheless , cancer cells should harbor some similarities to the cells from the organ of origin while being distinct to cells of other organs . We examined colonic adenoma that are initiated by stochastic loss of the second allele of the tumor suppressor gene Adenomatous Polyposis Coli ( Apc ) in our inducible , stem-cell–driven mouse model ( Lrig1CreERT2/+;Apcfl/+ ) [27 , 28] . We collected scRNA-seq data using inDrop for both the adenoma and adjacent normal ( Fig 6A ) and then appended sc-UniFrac analysis of these samples to the existing colon and pancreas analysis . As ranked by sc-UniFrac , the tumor landscape is dissimilar to the normal colon landscape , but it is more similar to the colon than the pancreas landscape , lying somewhere in the middle ( Fig 3D ) . This relationship can be approximated on principal component analysis ( PCA ) plots ( S8B Fig ) , in which global relationships between data points are better represented . Because all tumor cells would have activated Wnt signaling , there was minimal overlap between the tumor cell and normal cell landscapes , as expected . We performed experiments with control replicate samples to confirm minimal batch effects in this comparison . As such , adjacent normal colon was intermingled among normal colon samples , highlighting its normal phenotype . The distance-driving cells between tumor and adjacent normal were identified by branches with significant proportional shifts . For simplicity , we selected two subpopulations for further characterization: subpopulation 1 , which is the most skewed towards adjacent normal colon , and subpopulation 10 , which is composed entirely of cells from the tumor ( Fig 6B ) . Subpopulation 1 , when matched to the Mouse Cell Atlas , consists of differentiated absorptive cells of the gut ( Fig 6C ) . This result corroborates known colonic tumor biology that tumors are characterized by stem/progenitor signatures while the normal colon is overrepresented with differentiated cells . Note that the tumor sample also contributes to this subpopulation but only a very minor proportion , as tumors have differentiated cells at a very low level . Subpopulation 10 , which is completely tumor derived , represent granulocytes ( Fig 6D ) . Granulocytes—predominantly neutrophils—are absent in the uninflamed normal colon , while tumors present altered , possibly inflamed , microenvironments with substantial infiltrates . Next , we applied sc-UniFrac to analyze a scRNA-seq dataset describing oligodendrocyte progenitor cells ( OPCs ) that have been isolated from distinct regions of the mouse brain by microdissection [29] . While region-based information was provided , region-specific differences in OPC subpopulations were not identified in the original manuscript . We analyzed the cells from various brain regions together as in the original manuscript , with the assumption that technical variation between regions has been well-controlled for . Clustering by sc-UniFrac distance , we identified that OPCs grouped together on a dorsal ( cortex S1 , corpus callosum , hippocampus CA1 ) to ventral ( dentate gyrus , amygdala , zona incerta , striatum , hypothalamus , substantia nigra and ventral tegmental area [SN-VTA] , dorsal horn ) axis globally ( Fig 7A ) . Looking at more local clustering , we observed grouping of the dentate gyrus and amygdala OPCs , similar to previous work ( labeled as “immature” there ) , while SN-VTA and dorsal horn OPCs grouped together ( labeled as “mature” and also physically the most posterior regions of the central nervous system assayed ) ( Fig 7A ) . In addition , sc-UniFrac revealed that cortex S1 , corpus callosum , and hippocampus CA1 clustered together , while the zona incerta , striatum , and hypothalamus formed another cluster ( Fig 7A ) . These regions develop from pallium-derived and subpallium-derived tissues , respectively ( Fig 7B ) . The groupings by sc-UniFrac can be visually observed in t-SNE plots , supporting our analysis ( Fig 7C ) , although conclusions cannot be definitively drawn by visual inspection alone . Hence , sc-UniFrac was able to provide biologically meaningful results for relating OPCs from different regions of the brain . The presence of batch effects is a significant and common problem in scRNA-seq experiments , by introducing systematic error and masking underlying biological signals . Removal of batch effects is generally required prior to downstream analysis . Many methods and tools have been developed for batch correction [30–33] . Some methods have been successfully used in bulk RNA-seq [30 , 32] , while other methods were recently developed and specially designed for scRNA-seq [31 , 33] . While the suitability of batch-correction methods may depend on the distribution of data that vary from dataset to dataset , the universality of such methods is undefined given that there is no quantitative , objective metric to evaluate batch effect correction in scRNA-seq data . sc-UniFrac , a quantitative measure of cell population diversity in single-cell landscapes , provides a sensitive and objective way to assess the performance of batch-correction methods . We compared three batch removal methods , limma , ComBat , and MNN , on three scRNA-seq datasets . Limma and ComBat have been widely used for batch correction in bulk experiments , which fit a linear model to determine and then correct the batch effect for each gene [30 , 32] . MNN first identifies mutual nearest neighbor pairs between batches and then uses these pairs to estimate the batch effect in scRNA-seq data [34] . MNN is expected to perform well when population composition is different across batches . The evaluation of these methods was performed on the following three scRNA-seq datasets: ( 1 ) human embryonic kidney 293 ( HEK293 ) cells prepared fresh and cryopreserved from two batches [35] , ( 2 ) our three technical replicates of mouse colonic epithelium , and ( 3 ) two separate studies of mouse gastrulation [36 , 37] . For HEK293 cell line data , a small sc-UniFrac—reflecting high similarity—was observed between freshly isolated and cryopreserved samples within the same batch ( S9 Fig ) , indicating minimal technical variation during the cryopreservation process consistent with the original findings [35] . In contrast , a large sc-UniFrac was observed between two batches , indicating a strong batch effect similar to the original manuscript ( S9 Fig ) . All three methods , limma , ComBat and MNN , decreased the sc-UniFrac distance , indicative of batch effect correction ( Fig 8A and S10A–S10C Fig ) . Among them , limma and ComBat decreased the distances to those approaching to zero , suggesting that batch effects have been completely removed . In the technical replicates of the mouse colonic epithelium , only minimal batch effects were observed as indicated by our previous analyses . Due to the initial small differences between batches , limma , ComBat , and MNN only moderately removed batch effects further , as seen in the decrease in sc-UniFrac between replicates 1 and 2 , and between replicates 1 and 3 ( Fig 8B and S10D–S10F Fig ) . Batch effects were initially minimal between replicates 2 and 3 . In this case , limma and ComBat successfully removed the batch effects ( reduced sc-UniFrac to zero ) , while MNN failed to do so but instead introduced additional systematic bias ( sc-UniFrac increased ) ( Fig 8B ) . scRNA-seq data of mouse cells during gastrulation were obtained from two studies [36 , 37] , which used plate-based Smart-seq2 and G&T-seq ( genome and transcriptome sequencing ) , respectively , which introduced large technical variation . The first study generated scRNA-seq data from mouse embryos at E5 . 5 , E6 . 5 , and E6 . 75 , and the second focused on mouse embryos at E6 . 5 and E7 . 0 . sc-UniFrac generated from the uncorrected data indicated that the two datasets clustered by studies and not by developmental stages , revealing strong technical variation between the two studies ( Fig 8C and S11A Fig ) . After applying limma , ComBat , and MNN , sc-UniFrac indicated that the batch-correction methods removed the technical variation , with cells no longer clustering by studies but by developmental stages . Among the three methods , both limma and ComBat were able to arrange cells chronologically from the earliest development stage ( E5 . 5 ) to the latest ( E7 . 5 ) , whereas the ordering of samples processed by MNN was incomplete ( Fig 8C ) . This conclusion is supported by t-SNE analysis in which E6 . 5_1 and E . 6 . 5_2 clustered together after limma and ComBat but remained separated after MNN ( S11B–S11D Fig ) . From these results , limma and ComBat both outperformed MNN , probably due to identical population composition across batches in all datasets . One of MNN’s assumptions is that batch effects should be much smaller than biological variation , which may not hold true in these datasets . Additionally , the performance of MNN is dependent on the number of nearest neighbors to consider when identifying mutual nearest neighbor pairs . Choosing the correct parameter would probably improve MNN performance , but this would require prior knowledge of what the correct parameter is . While all methods were able to reduce technical variation , sc-UniFrac was able to quantitatively evaluate the initial batch effects and the performance of batch-correction methods .
We developed a new tool , sc-UniFrac , for quantitatively assessing the dissimilarities in cell population structures between two single-cell landscapes that can be generated by various types of single-cell technologies . Compared with existing methods , sc-UniFrac has distinct advantages , including ( 1 ) its ability to objectively and quantitatively assess population diversity differences , ( 2 ) its precision by taking population structure into account , ( 3 ) its statistical rigor based on the available UniFrac framework , ( 4 ) its intuitive and statistically robust method to identify disparate cell populations between samples , ( 5 ) its flexibility to analyze multiple samples and to add new samples to current analyses , and ( 6 ) its ability to handle any dataset with unlimited dimensional representation and distribution [16] . While sc-UniFrac presents a statistical test to estimate the significance of the observed distance , the resulting p-value should be interpreted with caution because it is sensitive to tree topology and randomization methods [38] . We have demonstrated the validity of sc-UniFrac using gold-standard datasets for which the similarities between datasets are known . Single-cell technologies provide unprecedented resolution to study heterogeneity in disease , especially in cancer . Intratumor heterogeneity is a key determinant of tumor diagnosis , prognosis , and drug response [39 , 40] . Although a large amount of effort has been devoted to the genomic [41] , transcriptomic [42–44] , and proteomic [45] subtyping of cancers in hope for better precision application and/or to better understand the disease , current bulk analyses obscure signals coming from distinct cell populations . Unbiased characterization of cellular diversity in tumor tissues and application of this information to define tumor subtypes provides a unique opportunity to better understand cancer . Subtypes based on tumor heterogeneity refine the subtypes defined by bulk “omic” approaches and may provide additional prognostic and diagnostic value for predicting patient survival and drug response . For current single-cell applications , comparing heterogeneity between multiple samples has been performed manually using t-SNE analysis in conjunction with distinguishing markers to qualitatively match cell populations across samples; however , this is done in a low-throughput fashion with few samples [46] . sc-UniFrac enables quantitative evaluation of cellular diversity among potentially large numbers of samples , which can then be rapidly clustered into different subtypes . Thus , sc-UniFrac can facilitate studies on intratumor and intertumor heterogeneity to reveal the importance of diverse cell populations in tumor progression and drug treatment . Furthermore , data structures generated by sc-UniFrac can be applied to software developed for microbiome research , such as QIIME [47] , which will provide single-cell researchers access to advanced analytical tools . Cell populations will expand , shrink , or emerge as a function of disease subtype , disease progression , or after extrinsic drug perturbations . sc-UniFrac can intuitively identify significantly altered cell populations or states driving compositional difference . Moreover , difference-driving cells can be further analyzed to identify gene expression signatures , and their identities and behaviors can be inferred based on transcriptomes of previously referenced cell types . Introduction of new cells into the landscape as a result of perturbation—e . g . , the infiltration of CD8 cytotoxic T cells into a tumor—can be deciphered by matching ( or blasting ) [48] difference-driving cells against the transcriptomes of reference cell types [3] , as sc-UniFrac has demonstrated here . There is currently a proliferation of single-cell data analysis tools , such that many of them utilize different approaches for achieving the same goal . In response , it is necessary for the single-cell biology community to benchmark the performance of these tools with reference datasets . Batch effect correction is a very important procedure for removing technical variation . The sources of variation can arise from runs on different sequencing lanes , different single-cell encapsulation platforms , ischemic times , or different tissue preparations , even if procedures are performed by the same experimenter . Several tools designed to remove batch effects have been developed specifically for scRNA-seq [21 , 34 , 49 , 50] . A quantitative measure of performance is required for effective benchmarking , and sc-UniFrac now provides a metric by which the similarity between single-cell landscapes of a tissue generated from different batches , before and after batch correction , can be evaluated . While evaluation of a tool on any specific datasets can be performed , it should be noted that the performance of any particular tool depends on the assumptions underlying the algorithm as well as the distribution of the dataset . Thus , different tools may perform better on some datasets than others . More importantly , this work calls attention to the requirement for proper experimental design and controls in scRNA-seq experiments . sc-UniFrac reports differences in single-cell landscapes agnostic to technical ( batch ) versus biological effects . Similar to other experimental platforms such as bulk RNAseq , if proper controls were not performed or an erroneous experimental design was adopted , technical effects can be confounded with biological effects to arrive at erroneous conclusions . Experimental samples should be prepared and run simultaneously with replicate control samples to assess whether the control samples produce similar landscapes , like we have done here with sc-UniFrac on technical and biological replicates . This is especially important for multi-sample experimental designs conducted over different batches . If technical effects were determined to overwhelm biological variation , appropriate measures—such as batch correction—should be performed prior to further downstream analysis . sc-UniFrac can support the evaluation of these effects , such that multiple single-cell landscapes clustering by batch or biological conditions can be readily evaluated in a quantitative and statistically robust manner . Furthermore , cell populations that contribute to batch effects can be identified through the sc-UniFrac pipeline , which helps delineate sources of technical variations , such as an excess of dying cells in one prep versus another . We have demonstrated various applications of this approach , and we envision its broad usage as increasing number of scRNA-seq datasets are generated .
Animal experiments were performed under protocols approved by the Vanderbilt University Animal Care and Use Committee and in accordance with NIH guidelines . Wild-type mice ( C57BL/6 ) and tumor-bearing mice ( Lrig1CreERT2/+;Apcfl/+ ) were euthanized in an approved fashion prior to dissection and tissue harvesting . Tumor induction was performed following a previous protocol [28] . sc-Unifrac is freely available as an R package at https://github . com/liuqivandy/scUnifrac . sc-Unifrac includes the following four main steps: Data processing: scRNA-seq data were first normalized by library size per cell ( total number of UMIs ) and log-transformed . Tree construction: highly variable genes were selected ( user defined: default 500 ) . PCA was performed to reduce the dimensionality while preserving the signal of interest , which reduces the noise and makes the data more tractable both from a statistical and computational point of view ( user defined: default = 4 ) . A hierarchical tree representing cell population structure was built by clustering via average linkage , and the upper portion of the tree was defined by cutting off the connections at k clusters ( user defined: default k = 10 ) . Quantification of cell population diversity: the sc-UniFrac distance was calculated by weighted branch sharing , and statistical significance was assessed by permutation testing . Identifying populations that drive sc-UniFrac by querying the shared branching structure: gene expression signatures were derived for matching against reference cell type signatures . The sc-Unifrac package provides the following two functions: ( 1 ) pairwise comparisons and ( 2 ) multi-sample comparisons . In pairwise comparison , sc-Unifrac generates a report to summarize the results , including the sc-UniFrac distance on population diversity , statistical significance , cell population structures , gene expression signatures in each altered population , and their match to reference cell types ( S1 Report ) . In multi-sample comparisons ( n samples ) , sc-Unifrac generates an n-by-n pairwise distance matrix , a corresponding statistical significance matrix , a hierarchical tree , and a table of counts per cluster per sample . sc-UniFrac ( D ) is calculated as: D=∑inbi×|AiAT−BiBT|∑jndj×|AjAT+BjBT| pvalue=|D*≥D|N Here , n is the total number of branches in the tree . bi is the length of branch i . Ai and Bi are the number of cells that descend from branch i in the two samples A and B , respectively . AT and BT are the total number of cells in two samples A and B , respectively . ∑jndj*|AjAT+BjBT| is the average distance of each cell from the root , used to normalized the distance from 0 to 1 . D* is the distance based on permuted data , while D is the observed distance . N is the total number of permutations . A distribution of distance is obtained with a p-value that reflects the probability that the permuted distances are greater than or equal to the observed distance by chance . Unshared branches are occupied by populations with statistically significant shifts between samples . The proportion shift of a cell population i is defined as psi=|AiAT−BiBT| , with statistical significance achieved if psi > ps* , where ps* is the proportion shift in permuted datasets . The signature composed of under- or overexpressed genes associated with a cell population was defined by limma comparisons with other populations . For predicting cell types , each cell of a given cluster was matched to the 894 cell type references from the Mouse Cell Atlas [3] . Matching was performed by deriving a Pearson correlation of all genes between the query cell and the reference cell type transcriptome . The p-Creode score was originally designed to compare p-Creode trajectories as a means to evaluate robustness of the calculated trajectories and also to arrive at a representative trajectory over multiple bootstrapped runs on the same single-cell dataset . Thus , it was developed to assess the dissimilarity between trajectories of largely the same sizes . Here , we use p-Creode scoring to compare different datasets , which can generate trajectories of different sizes . To make the p-Creode score size-invariant , we modified how nodes are transformed from one graph to another while maintaining the scoring approach outlined in [19] . Previously , when a node was not contained in both graphs , a node transformation was performed by translating the node in the second graph to the closest node in the first graph with a penalty ( S6A Fig ) . To eliminate excess penalty when a dense graph is transformed into a sparse graph and vice versa ( S6C Fig ) , the scoring routine was updated to allow for transformations into the edges as well as nodes in the reference first graph . More formally , the transformation penalty is the minimum distance to the closest edge projection between two nodes in opposing graphs plus the remaining graph edge distance to the closest node along the path of the pairwise comparison ( S6B Fig ) . The code and data for reproducing the analysis is at https://github . com/KenLauLab/pCreode_Comparison_Across_Datasets . We compared three batch-correction methods , limma , ComBat , and MNN . For limma , we used the removeBatchEffect function in the limma package , which fits a linear model to the data and then removes the component due to the batch effects [32] . For ComBat , we used the ComBat function in the sva package , which uses Empirical Bayes methods to adjust for both the mean and variance differences across the batches [30] . MNN identifies the mutual nearest neighbors between batches and uses them to estimate and remove the batch effect [34] . We performed the mnnCorrect function in the scran package . We set the number of nearest neighbors to consider to be 20 . Single-cell suspensions of colonic epithelium were prepared by chelating ( 3 mM EDTA; 1 mM DTT ) distal colon segments at 4°C for 45 minutes followed by shaking off crypts [19 , 51] . Isolated crypts were then dissociated into single cells using a DNase1/collagenase enzymatic cocktail ( 2 . 5 mg/mL DNase1 , 2 mg/mL collagenase ) at 37°C for 20 minutes . Crypt fragments were further mechanically dissociated into single cells using a 27 . 5-gauge needle . Cell suspensions were washed 2 times with cold PBS to remove debris and were enriched for live cells using a Miltenyi MACS dead cell removal kit . Live cell concentration was counted based on Trypan Blue positive cells , and a solution of 150 , 000 cells/mL was prepared for encapsulation . To maintain live cell viability , 18 ul of Optiprep was added per 100 ul of cell solution prior to encapsulation . Dissociation of pancreatic buds from E14 . 5 control embryos was performed using previously published protocols [52] . Briefly , pancreatic buds were dissected from control embryos and trypsinized followed by flow sorting . Single-cell suspensions from multiple embryonic buds were prepared , and a cell solution of 20 , 000 cells was prepared for encapsulation . Single-cell encapsulation was performed using the inDrop platform ( 1CellBio ) with an in vitro transcription library preparation protocol , as previously described [1] . inDrop utilizes CEL-Seq in preparation for sequencing and is summarized as follows: ( 1 ) reverse transcription ( RT ) , ( 2 ) ExoI nuclease digestion , ( 3 ) SPRI purification ( SPRIP ) , ( 4 ) single strand synthesis , ( 5 ) SPRIP , ( 6 ) T7 in vitro transcription linear amplification , ( 7 ) SPRIP , ( 8 ) RNA fragmentation , ( 9 ) SPRIP , ( 10 ) primer ligation , ( 11 ) RT , and ( 12 ) library enrichment PCR . Number of cells encapsulated was calculated by approximating the density of single-cell suspension multiplied by bead loading efficiency during the duration of encapsulation . Each sample was estimated to contain approximately 2 , 500 encapsulated cells . Following library preparation , as described above , the samples were sequenced using Nextseq 500 ( Illumina ) using a 150 bp paired-end sequencing kit in a customized sequencing run [19] . After sequencing , reads were filtered , sorted by their designated barcode , and aligned to the reference transcriptome using InDrops pipeline . Mapped reads were quantified into UMI-filtered counts per gene , and barcodes that correspond to cells were retrieved based on previously established methods [1] . From approximately 2 , 500 cells encapsulated , approximately 1 , 800 cells were retrieved per sample . Dissociation of colonic adenomas was performed in a two-phase process . In the first stage , colon adenomas were dissected from the distal colon and washed in ice-cold PBS . The tumors were digested in DMEM containing 2 mg/mL collagenase type II at 37 °C for 1 hour or until fragments had dispersed . The tumor tissue suspension was washed in ice-cold PBS and filtered through a 40 μm filter . Tumor epithelial crypts retained by the filter were collected and resuspended in PBS while the flow-through was discarded . The tumor epithelial fraction was filtered again through a 100 μm filter to remove undigested fragments , and the flow-through was collected . In the second stage , isolated tumor epithelial crypts were further digested into single cells for encapsulation similar to above . Technical replicates were different single-cell encapsulations collected from the same mouse colon but prepared and sequenced on different days . Biological replicates were tissues collected from different mice on different days but sequenced in the same run . sc-UniFrac was applied to two scRNA-seq datasets to test the time and memory cost , each composed of 25 , 507 genes and 1 , 000 cells . All tests were run without parallel computation on a Windows ( Microsoft; https://www . microsoft . com/en-us/ ) desktop with an Intel ( R ) Xeon ( R ) CPU E5-2620 0 at 2 GHz and 32 GB memory . With default parameters ( 500 highly variable genes , 4 PCs , and 10 clusters ) , sc-UniFrac took about 25 seconds to calculate the distance and statistical significance . The maximum memory used by sc-UniFrac was about 800 MB . The time and memory used only increased nominally with increasing numbers of clusters , increasing numbers of PCs , and increasing numbers of genes used . As expected , the running time increased linearly with the number of cells in each dataset ( Table 1 ) .
|
Single-cell technologies generate hundreds to thousands of data points per sample , presenting a conundrum in determining similarities and differences across multiple samples . Currently , similarity is determined by the degree of “intermixing” of data points among samples , a local approach , but this approach cannot accurately evaluate the similarity of samples with cell populations close in data space but not overlapping . We present sc-UniFrac , an approach to compare hierarchical trees that represent single-cell landscapes , taking both global and local similarities into account . Furthermore , sc-UniFrac allows cells that drive differences between samples to be easily identified as unbalanced branches on trees . We used sc-UniFrac to evaluate experimental design based on biological and technical replicates , regional specification of brain cells , degree and identity of stromal infiltrate into tumor , and computational batch-correction tools . sc-UniFrac will be an important analysis tool going forward as the cost of single-cell technologies drops and more studies adopt multi-sample experimental designs .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"blood",
"cells",
"infographics",
"medicine",
"and",
"health",
"sciences",
"ecology",
"and",
"environmental",
"sciences",
"methods",
"and",
"resources",
"immune",
"cells",
"cluster",
"analysis",
"statistics",
"hierarchical",
"clustering",
"immunology",
"mathematics",
"genome",
"analysis",
"cytotoxic",
"t",
"cells",
"information",
"technology",
"data",
"processing",
"digestive",
"system",
"research",
"and",
"analysis",
"methods",
"ecological",
"metrics",
"computer",
"and",
"information",
"sciences",
"white",
"blood",
"cells",
"genomics",
"animal",
"cells",
"mathematical",
"and",
"statistical",
"techniques",
"t",
"cells",
"species",
"diversity",
"gastrointestinal",
"tract",
"anatomy",
"ecology",
"data",
"visualization",
"cell",
"biology",
"graphs",
"transcriptome",
"analysis",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"computational",
"biology",
"colon",
"statistical",
"data"
] |
2018
|
Quantitative assessment of cell population diversity in single-cell landscapes
|
The establishment of HIV-1 latency gives rise to persistent chronic infection that requires life-long treatment . To reverse latency for viral eradiation , the HIV-1 Tat protein and its associated ELL2-containing Super Elongation Complexes ( ELL2-SECs ) are essential to activate HIV-1 transcription . Despite efforts to identify effective latency-reversing agents ( LRA ) , avenues for exposing latent HIV-1 remain inadequate , prompting the need to identify novel LRA targets . Here , by conducting a CRISPR interference-based screen to reiteratively enrich loss-of-function genotypes that increase HIV-1 transcription in latently infected CD4+ T cells , we have discovered a key role of the proteasome in maintaining viral latency . Downregulating or inhibiting the proteasome promotes Tat-transactivation in cell line models . Furthermore , the FDA-approved proteasome inhibitors bortezomib and carfilzomib strongly synergize with existing LRAs to reactivate HIV-1 in CD4+ T cells from antiretroviral therapy-suppressed individuals without inducing cell activation or proliferation . Mechanistically , downregulating/inhibiting the proteasome elevates the levels of ELL2 and ELL2-SECs to enable Tat-transactivation , indicating the proteasome-ELL2 axis as a key regulator of HIV-1 latency and promising target for therapeutic intervention .
Transcriptional silence of integrated HIV-1 proviruses in a minority of infected CD4+ T cells is a key signature of the latent viral reservoirs that necessitate a lifelong antiretroviral therapy ( ART ) to maintain their silence [1 , 2] . Strategies to expose the latently infected cells for immune recognition and clearance in individuals on ART rely on latency reversing agents ( LRAs ) to reactivate proviral transcription [3 , 4 , 5] . To date , multiple clinical trials have tested a variety of LRAs that are dominated by histone deacetylase ( HDAC ) inhibitors and NF-κB agonists [6] . However , only modest increases in viral transcription with little to no reservoir reduction are induced by these drugs [7] . Compared with the mechanisms used by the HDAC inhibitors and NF-κB agonists to relax chromatin and recruit RNA polymerase ( Pol ) II to the HIV-1 promoter , respectively , a less leveraged but arguably more specific and prominent feature of the HIV-1 transcriptional control is the Tat-dependent transition of Pol II from promoter-proximal pausing to productive elongation [8 , 9] . This rate-limiting step fuels a potent positive-feedback circuit to activate viral transcription without causing T cell activation [10] . Mechanistically , Tat stimulates HIV-1 transcriptional elongation by recruiting a specific host co-activator , the human Super Elongation Complex ( SEC ) [11 , 12] , to the paused Pol II through forming the Tat-SEC complex on the TAR RNA , a stem-loop structure located at the 5’ end of the nascent viral transcript [13 , 14] . The two critical components of the SEC , P-TEFb and ELL2 , stimulate Pol II elongation by different mechanisms and can thus synergistically induce the production of full-length viral transcripts [8] . In addition to residing in the SEC , P-TEFb also interacts with the bromodomain protein Brd4 , which competitively inhibits the Tat-SEC interaction [15 , 16] . The small molecule suppressor JQ1 binds to Brd4 to antagonize its inhibitory action on Tat-SEC , leading to the activation of HIV-1 transcription and latency reversal [17 , 18 , 19 , 20] . Notably , JQ1 is shown to synergize with other LRAs to reactivate latent HIV-1 in a number of ex vivo experiments involving CD4+ T cells from the ART-suppressed individuals [21 , 22 , 23 , 24] . A typical SEC contains P-TEFb , as well as one of each of the three pairs of homologous proteins: ELL1/ELL2 , AFF1/AFF4 , and ENL/AF9 . Owing to the ability of these proteins to create multiple different combinations among them , a family of related SEC complexes exists in cells . Our recent study shows that a low-abundance subset of SECs containing ELL2 and AFF1 play a predominant role in cooperating with Tat to reverse HIV-1 latency [25] . In fact , by simply increasing the cellular level of ELL2 , a highly unstable protein due to its polyubiquitination by the E3 ubiquitin ligase Siah1 and subsequent degradation by the proteasome [11 , 26 , 27] , it was possible to activate latent HIV-1 without using any drugs [25] . The complexity of the mechanisms that contribute to HIV-1 latency suggests that combinatorial LRAs from distinct mechanistic classes are necessary to expose the hidden viruses [3] . To identify novel classes of LRAs that cooperate well with the existing ones , we developed a genetic screen based on CRISPR interference ( CRISPRi ) [28] to look for additional host restriction factors that may represent previously unrecognized drug targets . By selecting authentic and effective CRISPRi targets through reiterative enrichments , we have identified several subunits of the proteasome as novel host factors that strongly inhibit HIV-1 transcription and promote latency . Our data indicate that the proteasome preferentially inhibits the Tat-dependent HIV-1 transcription by decreasing the cellular level of ELL2 , which in turn prevents formation of the ELL2-containing SECs . Furthermore , several FDA-approved proteasome inhibitors are shown to act synergistically with the existing LRAs to activate HIV-1 without inducing cell activation or proliferation in both cell line-based latency models and primary T cells from HIV-1-infected and ART-suppressed individuals . Collectively , our data indicate that targeting the proteasome-ELL2 axis provides a new avenue to expose the latent HIV-1 proviruses .
To identify novel human genes that inhibit HIV-1 expression , we set up a screen for the loss-of-function genotypes that could lead to the activation of latent HIV-1 provirus in the Jurkat-based 2D10 cell line , a widely used post-integration latency model with the d2EGFP-coding sequence in place of the viral nef gene in the proviral genome [29] . We first generated a 2D10-based TetOn mCherry-dCas9-KRAB cell line ( named 2D10-CRISPRi ) by adapting an inducible CRISPRi platform [30] . The loss-of-function genotypes were produced in this cell line by stably transducing a whole genome sgRNA library containing a total of ~200 , 000 sgRNAs at an average of 10 per gene [30] . Three days after the doxycycline ( Dox ) -induced production of the dCas9-KRAB fusion , the cells were subjected to fluorescence-activated cell sorting ( FACS ) to isolate the GFP+ cells containing activated HIV-1 ( Fig 1A ) . Because the first round of selection did not yield any positive signals that were above the background ( Fig 1B ) , we decided to repeat the procedure a few more times in the hope of enriching the desired genotypes ( Fig 1A ) . To do this , the sgRNA sequences were PCR-amplified from the genome of GFP+ cells isolated from the previous round of FACS and cloned into the empty vector to generate an enriched sgRNA library , which was then transduced into the original 2D10-CRISPRi cells for the next round of selection ( Fig 1A ) . This procedure , called the Reiterative Enrichment and Authentication of CRISPRi Targets or REACT , was repeated 4 times . As expected , the CRISPRi-induced HIV-1 activation began to noticeably increase in the cell population starting from round 3 ( 1 . 06% GFP+ cells under Dox+ vs . 0 . 46% spontaneous reactivation under Dox- conditions ) and culminating in round 5 of REACT ( 4 . 27% vs . 0 . 55%; Fig 1B ) . High-throughput sequencings of the sgRNA libraries enriched by rounds 1 , 2 , 3 , and 4 demonstrate a progressive enrichment of several main hits from background noise ( Fig 1C and S1 Fig ) . The 7 most significantly enriched sgRNAs target 6 different genes: PSMD1 , NFKBIA , CYLD , GON4L , PSMD3 , and PSMD8 ( Fig 1D ) . Among these , NFKBIA ( aka . IκBα , an inhibitor of NF-κB ) and CYLD ( a deubiquitinase for NFKBIA ) have been reported to encode suppressors of HIV-1 transcription [31 , 32] . It is thus unsurprising that their silencing caused viral activation . On the other hand , PSMD1 , PSMD3 , and PSMD8 , all encoding the canonical proteasome subunits , and GON4L , which encodes a protein found in a transcriptional co-repressor complex with HDAC1 [33] , represent previously unreported and potentially novel host restriction factors for HIV-1 . To verify that the 6 genes identified by REACT indeed encode the restriction factors that promote HIV-1 latency , we synthesized and cloned the top 7 sgRNA hits and a negative control sequence into the empty vector used to generate the library , and stably transduced them into the 2D10-CRISPRi cells . The RT-qPCR and FACS analyses indicate that the 7 sgRNAs but not the negative control downregulated the expression of their respective target genes ( Fig 2A ) and efficiently activated HIV-1 ( Fig 2B ) . Further RT-qPCR analyses demonstrated that after the 6 genes were downregulated by CRISPRi , the HIV-1 env mRNA level increased by one to two orders of magnitude ( Fig 2C ) , whereas the cellular GAPDH transcript remained mostly unchanged ( Fig 2D ) . To further confirm that the proteasomal subunits can be downregulated to activate latent HIV-1 , we used siRNAs to knock down the expression of PSMD1 , PSMD3 , as well as a non-proteasomal REACT target CYLD in two different HIV-1 latency model cell lines , Jurkat 2D10 [29] and J-Lat A2 , which contains an integrated , transcriptionally silent LTR-Tat-Flag-IRES-EGFP cassette [34] . Like the CRISPRi-induced silencing , the knockdown ( KD ) by RNA interference ( RNAi ) significantly reactivated latent HIV-1 and enhanced mRNA production from the HIV-1 LTR but not the GAPDH promoter in both systems ( S2A–S2F Fig ) . Transcriptional silencing leads to HIV-1 latency . It is thus important to determine whether the 6 genes identified by REACT directly affect HIV-1 transcription . Specifically , we asked whether they influence Tat-transactivation , which is the most prominent feature of HIV-1 transcription . To this end , we examined the impact of the CRISPRi-induced downregulation of the 6 genes on expression of an integrated , HIV-1 LTR-driven luciferase reporter gene in Jurkat-based 1G5 [35] and 1G5+Tat cell lines [36] . The data indicate that downregulating PSMD1 , PSMD3 , PSMD8 , NFKBIA and CYLD significantly increased the LTR-driven luciferase expression only in 1G5+Tat cells that constitutively express Tat ( Fig 2E & 2F ) . In contrast , targeting GON4L by two different sgRNAs did not increase the LTR activity in either cell line . Together , these data implicate the proteasomal subunits PSMD1 , PSMD3 , and PSMD8 as novel host factors that inhibit Tat-dependent HIV-1 transcription and promote viral latency . Since PSMD1 , PSMD3 and PSMD8 are all located in the 19S regulatory particle of the 26S proteasome [37] , we asked whether subunits in the 20S core particle also restrict HIV-1 activation . To answer this question , we used shRNAs to knock down two core subunits , PSMA1 and PSMB1 , in 2D10 cells and discovered that the KD potently reactivated latent HIV-1 and increased the viral env but not cellular GAPDH mRNA level ( Fig 3A–3C ) . Considering these results , we further tested whether inhibiting the proteasomal function with drugs could also reactivate latent HIV-1 . We treated 2D10 and A2 cells with MG132 , which is frequently used in research settings , three FDA approved proteasome inhibitors: Bortezomib ( Millennium , Velcade , PS-341 ) , Carfilzomib ( ONYX , PR171 ) , Ixazomib ( Millennium , MLN2238 ) , as well as three inhibitors in late stage clinical trials: Ixazomib citrate ( MLN9708 ) , Oprozomib ( ONYX , ONX0912 ) , Delanzomib ( CEP18770 ) . The results show that when used at nano- to submicro-molar concentrations , all the inhibitors were able to dose-dependently increase the HIV-1 LTR-driven transcription and reverse viral latency in up to ~80% of 2D10 and ~30% of A2 cells ( Fig 3D–3I ) . Notably , these drug concentrations only mildly affected cell viability and no more than 50% cell death was observed even under the highest concentrations used ( S3A–S3D Fig ) . Consistently , downregulating individual proteasome subunits by CRISPRi or shRNA for 3–5 days was in general fairly tolerated , although the loss of PSMD1 or PSMD3 produced a more prominent effect on cell viability ( 45% and 59% viable cells , respectively ) compared to the loss of PSMD8 , PSMA1 or PSMB1 ( 73% , 92% and 91% , respectively; S3E & S3F Fig ) . Together , our data indicate that targeting the proteasome by either gene silencing or drugs can effectively promote HIV-1 transcription and latency reversal . To investigate the impact of inhibiting the proteasome in a more clinically relevant setting , we selected the two FDA-approved proteasome inhibitors , bortezomib and carfilzomib , and assessed their abilities to reactivate latent HIV-1 in CD4+ T cells isolated from 11 HIV-1-infected individuals on suppressive ART ( S1 Table ) . While 10 nM bortezomib alone was able to reactivate HIV-1 by ~2-fold , the more significant finding is that both inhibitors potently enhanced the latency-reversing effects of existing LRAs at concentrations ( 10–100 nM; Fig 4A ) that were effective in anti-cancer treatments [38 , 39] . For example , when used alone , 1 μM JQ1 , a BET bromodomain inhibitor and known LRA [19 , 23] , reactivated HIV-1 just 4-fold compared to the DMSO control . However , bortezomib and carfilzomib enhanced this effect up to 19-fold ( Fig 4A ) . Furthermore , 40 nM of the HDAC inhibitor romidepsin [40 , 41] reactivated HIV-1 5-fold by itself , but produced up to 27-fold activation together with bortezomib or carfilzomib ( Fig 4A ) . Finally , while 1 μM of the HDAC inhibitor vorinostat ( aka SAHA; [42] ) alone did not activate HIV-1 in a statistically significant manner , its combination with bortezomib or carfilzomib caused up to 11-fold activation ( S4 Fig ) . Notably , the PKC agonist bryostatin [43] produced no obvious effect either alone or together with the two proteasome inhibitors ( S4 Fig ) . Using the Bliss Independence model for assessing drug synergism [21 , 24 , 44] , we discovered that the co-administration of bortezomib ( both 10 and 100 nM ) or carfilzomib ( 10 nM ) with JQ1 ( 1 μM ) or romidepsin ( 40 nM ) all exhibited robust synergistic activity ( p<0 . 05 ) ( Fig 4B ) . In addition , after 24 hours of treatment at 10 and 100 nM concentrations , which were the same conditions used in the latency-reversal assay , bortezomib and carfilzomib did not induce the surface expression of CD25 and only marginally induced CD69 in cells from 3 patients ( Fig 4C & S5 Fig ) . This contrasts with the robust induction of the two activation markers by PMA plus ionomycin as well as the considerable CD69 induction by vorinostat ( Fig 4C & S5 Fig ) . Furthermore , staining with CellTrace CFSE detected no proliferation of live primary CD4+ T cells after the initial 24-hour exposure to the proteasome inhibitors and then additional 3 days of culture in the absence of the drugs ( Fig 4D & S6 Fig ) . Finally , at least during the initial 24-hour treatment , the two inhibitors were well-tolerated by cells from all 3 patients , with the cells from patient #1 showing only a mild loss of viability in the presence of 10 nM carfilzomib for up to 4 days ( S7 Fig ) . In summary , these data demonstrate that the proteasome inhibitors can synergize with existing LRAs to potently reactivate HIV-1 ex vivo without inducing activation or proliferation of the patient-derived primary CD4+ T cells . Consistent with the CRISPRi result in Fig 2E , the bortezomib inhibition of the proteasome also enhanced the HIV-1 LTR-driven transcription in a Tat-dependent manner ( Fig 5A ) . HIV-1 transcriptional elongation , especially the Tat-activated process , is exquisitely controlled by a network of P-TEFb complexes that include the 7SK snRNP , the SECs and the Brd4-P-TEFb complex [45] . In light of this revelation , we examined whether the levels of P-TEFb , its major known associated factors as well as the NFκB-inhibitor IκBα , which is believed to be regulated at the protein stability level [46 , 47] , would change after the downregulation of the proteasome . Examination of cell extracts by Western blotting demonstrates that among all the proteins analyzed , downregulating the proteasome in Jurkat cells by CRISPRi against PSMD1 , PSMD3 , and PSMD8 ( Fig 5B ) , or RNAi against PSMA1 and PSMB1 ( Fig 5C ) consistently elevated the protein levels of only ELL2 and occasionally ELL1 ( e . g . after CRISPRi against PSMD1 & PSMD3 ) , which are two alternative subunits of the SECs [48] . Notably , the mRNA level of ELL2 was not elevated , but the ELL1 mRNA level was somewhat increased in this process ( S8 Fig ) . The elevated ELL2 protein level as a result of the proteasomal downregulation is consistent with the previous reports showing that ELL2 is tightly controlled by the E3 ubiquitin ligase Siah1-induced degradation by the proteasome [11 , 26 , 27] . Of note , inhibiting the proteasome by bortezomib also elevated the ELL2 protein level in Jurkat nuclear extract , which in turn resulted in the formation in the nuclei of more ELL2-containing SECs as revealed by anti-CDK9 immunoprecipitation followed by Western blotting ( Fig 5D ) . Since among all the related members of the family of SEC complexes , the ELL2-containg SECs play a predominant role in supporting Tat-transactivation and reversing viral latency [25] , we compared the bortezomib-induced HIV-1 activation in three different 2D10-based cell lines: WT [29] , ΔELL2 ( ELL2-knockout ) and ΔELL2-R2 ( ΔELL2 cells containing an integrated vector expressing ELL2-Flag to approximately the endogenous level ) [25] . The FACS analysis demonstrates that compared to WT 2D10 cells , the absence of ELL2 in ΔELL2 cells abolished the bortezomib-induced HIV-1 latency reversal , which was efficiently rescued by expressing ELL2-Flag in the ΔELL2-R2 cells ( Fig 5E ) . Taken together , these results indicate the stabilization of ELL2 and elevated formation of the ELL2-SECs as a key mechanism for promoting HIV-1 Tat-transactivation and latency reversal in CD4+ T cells upon the inhibition/downregulation of the proteasome ( Fig 5F ) .
In this study , we have developed a CRISPRi-based screen to reiteratively enrich loss-of-function genotypes that promote HIV-1 transcription in latently infected CD4+ T cells . The identified hits include the not-so-surprising factors that suppress the NF-κB pathway ( NFKBIA , CYLD ) or interact with the HDAC complex ( GON4L ) , as well as three unexpected proteasomal subunits . Our subsequent experiments employing RNAi to target these three and also two other core subunits of the proteasome and testing various proteasome inhibitors in two different cell line-based latency models as well as primary CD4+ T cells from HIV-infected individuals on suppressive ART all support the notion that targeting the proteasome is an effective strategy to expose latent HIV-1 . Interestingly , a study published in 2004 has shown that the mRNA levels of multiple genes encoding the various proteasome subunits are upregulated in latently-infected cell lines and that treating these cells with a proteasome inhibitor CLBL stimulated lytic viral replication [49] . Based on these early revelations and our current study , which employs multiple proteasome inhibitors and extends the analysis to primary CD4+ T cells , we propose that the elevated proteasome level in HIV-infected cells is a key mechanism used to silence viral transcription and drive the virus into latency . Consistent with a previous report showing that the proteasome inhibitors can enhance the P-TEFb-mediated HIV-1 transcriptional elongation [50] , our current study pinpoints ELL2 , which joins P-TEFb , AFF1 and ENL/AF9 to form the ELL2-SECs especially important for Tat-transactivation [11 , 25] , as the target of the proteasome inhibitors . This insight as well as the observation that the Tat-dependent HIV-1 transcription is preferentially affected by targeting the proteasome ( Figs 2E , 2F & 5A ) allow us to propose a model in Fig 5F . According to this model , in latently infected cells , the elevated proteasome level keeps the ELL2 concentration low through polyubiquitination and proteasomal degradation [26] . This prevents the assembly of the ELL2-SECs and blocks HIV-1 transcription . Upon downregulating/inhibiting the proteasome , the blockage is removed to increase the cellular ELL2 level . This results in the formation of more ELL2-SECs to stimulate Tat-transactivation , which in turn generates more Tat to fuel a robust positive feedback loop for HIV to exit latency . We have recently shown that the poly-ADP-ribosylation enzyme PARP1 upregulates ELL2 expression through inhibiting transcription as well as inducing degradation of Siah1 [27] , the demonstrated E3 ubiquitin ligase for ELL2 [26] . During the control of HIV-1 transcription , the PARP1-Siah1 axis and the proteasome display strong similarities: Both preferentially affect the Tat-dependent transactivation process , and both accomplish this by controlling the cellular levels of ELL2 and ELL2-SECs . Because the PARP1-Siah1 axis works upstream of the proteasome-dependent regulation of ELL2 [27] , it is tempting to speculate that by simultaneously augmenting PARP1 function and inhibiting the proteasome , it is possible to synergistically reactivate latent HIV-1 , a hypothesis that is worth testing in future studies in primary CD4+ T cells . The proteasome has been extensively characterized as a therapeutic target for treating both hematologic and solid tumors; and a number of inhibitors have been developed and approved for this purpose [51 , 52] . Our present study indicates that in addition to their anti-cancer effects , the two FDA-approved proteasome inhibitors , bortezomib and carfilzomib , can also synergize with existing LRAs such as JQ1 and romidepsin to reverse HIV latency in resting CD4+ T cells from ART-suppressed individuals without inducing T-cell activation or proliferation ( Fig 5 ) . Future studies will inform us whether this effect can also be detected in real clinical settings involving HIV patients . Moreover , the safety and efficacy of combining the proteasome inhibitors with other LRAs to expose the latent HIV-1 reservoirs for eradication also await further evaluation . It is known that the proteasome regulates CD4+ T cell activation and proliferation through controlling cellular levels of various cyclins and cyclin-dependent kinase inhibitors , and that inhibiting proteasomal activity suppresses essential functions of activated CD4+ T cells [53 , 54] . In addition , the proteasome also modulates fate specification of CD8+ T cells during differentiation . Inhibiting the proteasome increased the number of effector CD8+ T cells and reduced the proportion of memory CD8+ T cells , and the inhibitor-treated CD8+ T cells exhibited increased killing of target cells in cytotoxicity assays [55 , 56] . Thus , proteasome inhibitors may suppress undesired CD4+ T-cell activation induced by other LRAs in HIV-infected individuals and promote killing of infected cells by CD8+ cells at the same time . Future studies will be needed to investigate the immunologic ramifications of proteasome inhibition in HIV-infected individuals . Methodologically , the REACT protocol described here represents a significantly improved strategy to identify authentic genotypes that are hidden in a noisy background . Due to the stochastic nature of HIV-1 transactivation [57 , 58] , the GFP-based HIV-1 latency models always display a small percentage of GFP-positive cells due to a low level of spontaneous viral activation [29 , 34] . This background noise could potentially mask and overwhelm the real signals in any genome-wide screens that must start with a pooled library . The complexity of such libraries causes each genotype to have an extremely low representation in the whole population . Therefore , the phenotypic change induced by a to-be-identified genotype in only a few cells , even though genuine and significant , could easily be lost in a noisy background as exemplified by the first two rounds of REACT in our study . Only through repeated cycles of enrichment , the desired genotypes can be progressively enriched and become prominent in the population as demonstrated by high throughput sequencing of the sgRNA libraries enriched from round 1 to 4 of REACT ( Fig 1B and S1 Fig ) . Thus , although REACT may under-sample genotypes that inhibit cell growth , it can still be very useful for identifying the genetic basis of other noisy phenotypes that are not amenable to the single-round genome-wide screens .
The part of this study utilizing specimens from HIV-infected individuals was approved by the UCSF Committee on Human Research . All research participants were recruited from the UCSF SCOPE cohort after obtaining written informed consent , and all subject data and specimens were coded to protect confidentiality . All participants were adults and met strict selection criteria and had well-documented persistent viral suppression for over 7 years ( S1 Table ) . All the Jurkat-based cells were maintained in RPMI 1640 medium with L-glutamine , 10% fetal bovine serum ( FBS ) , 100 IU/ml penicillin , and 100 μg/ml streptomycin in 5% CO2 at 37°C . To prepare the CRISPR interference ( CRISPRi ) platform , Jurkat 2D10 cells ( previously generated by Karn lab based on human CD4+ T cells Jurkat line [29] ) were transduced by pHR-TRE3G-Krab-dCas9-P2A-mCherry [30] and pLVX-advanced-TetOn ( from IGI , UC Berkeley ) . A clone ( named “2D10-CRISPRi” ) expressing KRAB-dCas9-HA-P2A-mCherry in response to doxycycline ( Dox ) was first selected by fluorescence-activated cell sorting ( FACS ) and then verified by Western blotting [59] . To start REACT , a genome-wide CRISPRi sgRNA library in pSico-based vector with BFP marker and puromycin-resistance [30] were packaged using a 3rd generation lentiviral packaging system and transduced into 2×108 2D10-CRISPRi cells at an efficiency of ~40% . Two days after transduction , the non-transduced cells were killed by adding puromycin into the medium to a final concentration of 1 μg/ml for 3 days , at which time more than 95% of the surviving cells were BFP-positive as confirmed by FACS . About 5×107 of the cells were pooled and treated with 1 μg/ml Dox for 3 days in 400 ml medium . The Dox-treated cells were then selected by FACS for the GFP/mCherry/BFP triple-positive phenotype . Then sgRNA cassettes were PCR-amplified from the genomes of the selected cells using the primer pair: REACT-5F ( 5’-GCACAAAAGGAAACTCACCCT-3’ ) / REACT-3R ( 5’-CGACTCGGTGCCACTTTTTC-3’ ) . After digestion with BstXI and BlpI , the cassettes were cloned into the empty library vector pSico-BFP-puro [30] and then amplified in E . coli and extracted as an enriched library , which was then transduced into the original 2D10-CRISPRi cells for the next round of REACT . Upon repeating the procedure 4 times , the sgRNA sequences from the enriched libraries and original library were amplified by using the index primer pairs CRISPRi_TSS_12_P5/CRISPRi_TSS_12_P7 or CRISPRi_TSS_6_P5/CRISPRi TSS_6_P7 , and deep-sequenced by using the primer CRISPRi TSS_seq V2 . The sequences of the primers were listed in S2 Table . The deep-sequencing results were then converted into sgRNA counts by using the ScreenProcessing tool [60] . The fold of enrichment for each sgRNA sequence was calculated based on its reads per million in the round 4-enriched library divided by those in the original library and presented on a scatter plot . The Jurkat-based HIV-1 latency models 2D10 ( previously generated by Karn lab based on human CD4+ T cells Jurkat line [29] ) and J-Lat A2 ( previously generated by Verdin lab based on human CD4+ T cells Jurkat line [34] ) were first treated in triplicates with 0 . 1% DMSO , 1 μg/ml Dox or the various concentrations of proteasome inhibitors , and then re-suspended in cold phosphate-buffered saline ( PBS ) . Quantification of the GFP+ cells was conducted on a BD Bioscience LSR Fortessa X20 cytometer . The data were analyzed with the Flowjo software and plotted as bar graphs . The preparations of the CRISPRi platform in Jurkat 1G5 [35] and 1G5+Tat cells [36] ( both kind gifts from Dr . Melanie Ott in the Gladstone Institutes , San Francisco ) were the same as in 2D10 cells described above . DNA oligos containing the sgRNA sequences identified by REACT and a negative control ( 5’-GCAGCATGCTCGCCTGGCTGC-3’ ) were synthesized and cloned into the pSico-BFP-puro vector and stably transduced into the 1G5-/+Tat-CRISPRi cells . For the luciferase assay , 1x106 of the cells were treated with 0 . 1% DMSO or 1 μg/ml Dox in triplicates for three days , and then lysed in 200 μl 1 × Reporter Lysis Buffer ( Promega ) , frozen-thawed once , and centrifuged at 20 , 800 × g for 1 min at 4 °C . Luciferase activities in the supernatants were measured by using the Luciferase Assay System ( Promega ) on a Lumat LB 9501 luminometer . The relative luciferase units in the Dox-treated cells were normalized to those of the DMSO-treated cells and plotted as bar graphs . Total cellular RNAs were extracted by using the RNeasy kit ( Qiagen ) and reverse transcribed by M-MLV Reverse Transcriptase ( VWR , M1701 ) with random hexamers ( Invitrogen , 48190–011 ) . The cDNAs were subjected to qPCR using DyNAmo HS SYBR Green qPCR kit ( Fisher , F-410L ) on a CFX96 system ( Bio-Rad ) with the primer pairs listed in S2 Table . All reactions were carried out in triplicates . The PCR signals were normalized to those of ActB and displayed as bar graphs . Jurkat 2D10 cells were transduced with the pLKO . 1-puro lentiviral vectors containing shScramble ( 5’-CCTAAGGTTAAGTCGCCCTCG-3’ ) , shPSMA1 ( 5'-AATGATTATACAGACCCTTTC-3' ) , or shPSMB1 ( 5’-AAGACATCTTTCACCAGCCGC-3’ ) sequences . Two days after transduction , puromycin was added to the medium to a final concentration of 1 μg/ml to obtain and maintain the stable KD pools . The KD efficiencies of the pools were examined by RT-qPCR and Western blot . Jurkat 2D10 or J-Lat A2 cells were aliquoted into 24-well plates in 1 ml medium at the density of 5×105 cells/ml . Each of the following drugs was added to the cells in triplicate at the indicated concentrations: MG132 ( Sigma , M7449 ) , Bortezomib ( Millennium , Velcade , PS-341 , from Selleckchem , S1013 ) , Carfilzomib ( ONYX , PR171 , from Selleckchem , S2853 ) , Ixazomib ( Millennium , MLN2238 , from Selleckchem , S2180 ) , Ixazomib citrate ( MLN9708 , from Selleckchem , S2181 ) , Oprozomib ( ONYX , ONX0912 , from Selleckchem , S7049 ) , and Delanzomib ( CEP18770 , from Selleckchem , S1157 ) . For control groups , 0 . 1% DMSO was used . After 16~20 hours treatment , the cells were subjected to FACS to measure the percentages of GFP+ cells as described above . Cell viabilities were determined by Forward Scatter vs . Side Scatter gating using untreated cells as the control . Approximately 2 × 108 2D10 cells maintained in 400 ml RPMI medium were treated with 10 nM bortezomib or 0 . 1% DMSO for 18 hours . The following operations were all carried out at 4 °C or on ice unless stated otherwise . The cells were harvested and swollen in 4 ml hypotonic buffer A [10 mM HEPES-KOH ( pH 7 . 9 ) , 1 . 5 mM MgCl2 and 10 mM KCl] for 5 min and then centrifuged at 362 g for 5 min . The cells were then disrupted by grinding 20 times with a Dounce tissue homogenizer ( Fisher K8853000007 ) in 4 ml buffer A , followed by centrifugation at 3 , 220 g for 10 min to collect the nuclei . The nuclei were then extracted in 450 μl buffer C [20 mM HEPES-KOH ( pH 7 . 9 ) , 0 . 42 M NaCl , 25% glycerol , 0 . 2 mM EDTA , 1 . 5 mM MgCl2 , 0 . 4% NP-40 , 1 mM dithiothreitol and 1 × protease inhibitor cocktail] for 30 min and then subjected to centrifugation at 20 , 800g for 10 min . The supernatant ( nuclear extracts or NE ) was mixed with 4 μg anti-CDK9 antibody [11] for 45 min , and then with 15 μl Protein A agarose ( Invitrogen 15918–014 ) for 1 hr with rotation . After being washed three times with 1 ml buffer D [20 mM HEPES-KOH ( pH 7 . 9 ) , 0 . 3 M KCl , 15% glycerol , 0 . 2 mM EDTA and 0 . 4% NP-40] , the beads were eluted with 30 μl 0 . 1 M glycine-HCl ( pH 2 . 0 ) at room temperature for 15 min . For western blot , 3% of the NE input and 50% of the immunoprecipitation eluate from each treatment condition were analyzed . The primary antibodies used for Western blots are listed in S3 Table . The primary antibodies were diluted to 1 μg/ml and the secondary antibodies were diluted 10 , 000-fold . Fresh blood ( 100 ml ) was collected and peripheral blood mononuclear cells ( PBMCs ) were isolated from whole blood using Lymphocyte Separation Medium ( Corning 25-072-CI ) . CD4+ T cells were isolated from PBMCs using negative selection by EasySep kit ( STEMCELL 19052 ) according to the manufacturer’s instructions . Isolated CD4+ T cells were aliquoted at a density of 1×106 cells per well in 1 mL RPMI medium plus 10% FBS in a 48-well flat-bottom plate . The cells were treated with 0 . 2% DMSO , 25 μl ( 1:1 bead:cell ratio ) αCD3 + αCD28-conjugated beads ( Dynal 11131D ) , 50 ng/ml ( 81 nM ) PMA + 1 μM Ionomycin , 1 μM vorinostat , 1 μM JQ1 , 40 nM romidepsin , or 10 nM bryostatin alone or in combination with 10 nM or 100 nM bortezomib or carfilzomib for 24 hr . All drugs were prepared in the culture medium from stock solutions dissolved in DMSO . After the treatment , total RNAs from the cells were extracted by 1 ml TRIzol Reagent ( Invitrogen 15596026 ) . The RNAs were reverse-transcribed using M-MLV Reverse Transcriptase ( VWR , M1701 ) with random hexamers ( Invitrogen , 48190–011 ) . The HIV-1 RNA was quantified by qPCR using DyNAmo HS SYBR Green qPCR kit ( Fisher , F-410L ) with the HIV-1-specific primers F522-43 ( 5’-GCCTCAATAAAGCTTGCCTTGA-3’ ) and R626-43 ( 5’-GGGCGCCACTGCTAGAGA-3’ ) [24] on a CFX96 system ( Bio-Rad ) . All reactions were carried out in triplicates . The PCR signal from each drug combination was normalized to the DMSO group for each individual to calculate the fold induction and displayed in scatter plots . We adapted the Bliss independence model [44] as implemented by previous studies [21 , 22 , 24] to test for synergy when different concentrations of bortezomib and carfilzomib were combined with JQ1 or romidepsin ex vivo . For drugs x and y , we used the equations faxyP = fax + fay— ( fax ) ( fay ) , and Δfaxy = faxyO—faxyP . Here , fax and fay represent the observed effects of drug x and drug y respectively , faxyP represents the predicted fraction affected by the combination of drug x and drug y if there is no synergy or antagonism between drug x and drug y; faxyO represents the observed effect when x and y were tested together . Calculation of fax utilized the following approach adapted from the above cited publications: fax = ( HIV RNA copies with drug x—background copies with DMSO ) / ( HIV RNA copies with αCD3-αCD28 stimulation—background copies with DMSO ) . The copy number of HIV RNA was determined by extrapolation against a 7-point standard curve ( 1–1 , 000 , 000 copies ) prepared from a synthetic HIV cDNA fragment . In cases where one or more experimental drug conditions resulted in RNA expression exceeding the αCD3-αCD28 stimulation , we imputed the highest HIV RNA value in that experiment +1 to represent the denominator for calculation of fax . Statistical significance was calculated by two-tailed Student’s t-test comparing faxyO and faxyP ( *: p < 0 . 05 , **: p < 0 . 01 , and ***: p < 0 . 001 ) . With this model , Δfaxy > 0 with statistical significance ( p < 0 . 05 ) indicates synergy , Δfaxy = 0 indicates additive effect ( Bliss independence ) , Δfaxy < 0 with statistical significance indicates antagonism . CD4+ T cells isolated from HIV-infected ART-suppressed individuals were treated with 0 . 2% DMSO , 50 ng/ml ( 81 nM ) PMA and 1 μM Ionomycin , 1 μM vorinostat , 10 nM bortezomib , 100 nM bortezomib , 10 nM carfilzomib , or 100 nM carfilzomib for 24 hours . The cells were stained with LIVE/DEAD Cell Stain Kit ( Invitrogen , L34955 ) , and then stained with PE-conjugated mouse anti-Human CD69 antibody ( BD Biosciences , 555531 ) , and FITC-conjugated mouse anti-Human CD25 antibody ( BD Biosciences , 555431 ) . Flow cytometry was conducted on a BD Bioscience LSR Fortessa X20 cytometer , and data were analyzed using the Flowjo software . CD4+ T cells isolated from HIV-infected ART-suppressed individuals were stained with 10 μM 5 ( 6 ) -carboxyfluorescein N-hydroxysuccinimidyl ester ( CFSE , Abcam ab113853 ) for 15 min . The cells were treated subsequently with 0 . 2% DMSO , 10 nM bortezomib , 100 nM bortezomib , 10 nM carfilzomib , or 100 nM carfilzomib for 24 hr , and then washed and cultured for another 3 days in fresh medium . The cells were stained with LIVE/DEAD Cell Stain Kit ( Invitrogen , L34955 ) and PE-conjugated anti-human CD4 Antibody ( BioLegend , 317410 ) . Flow cytometry and data analysis were conducted as described above . CD4+ T cells isolated from HIV-infected ART-suppressed individuals were treated with the various drugs for 4 days as described above . On day 1 , 2 , 3 , and 4 , an aliquot of cells from each treatment was stained with LIVE/DEAD Cell Stain Kit ( Invitrogen , L34955 ) . Untreated cells were used for day 0 . Flow cytometry and data analysis were conducted as described above .
|
To cure chronic HIV-1 infection requires reversal of HIV-1 latency from latently infected CD4+ T cells . A key step in HIV latency reversal is the recruitment of Super Elongation Complexes ( SECs ) that contain ELL2 by an HIV-encoded protein , Tat , to activate proviral transcription . To identify novel drug targets , we conducted a CRISPRi-based screen to enrich the sgRNAs that increase HIV transcription in latently infected CD4+ T cells . Three of the six most prominent hits in our screen are proteasome subunits . We further proved that antagonizing the proteasome promotes Tat-induced HIV-1 transcription in cell line-based latency models . Furthermore , we found that two FDA-approved proteasome inhibitors strongly synergize with existing LRAs ex vivo without inducing cell activation or proliferation . We further found that antagonizing the proteasome elevates the levels of ELL2 and ELL2-containing SECs in the cells , thus enabling Tat-transactivation . These results indicate that the proteasome-ELL2 axis is a key regulator of HIV-1 latency could potentially be targeted for therapeutic interventions .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"blood",
"cells",
"flow",
"cytometry",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"cancer",
"treatment",
"immunology",
"microbiology",
"retroviruses",
"viruses",
"immunodeficiency",
"viruses",
"oncology",
"rna",
"viruses",
"proteasome",
"inhibitors",
"research",
"and",
"analysis",
"methods",
"specimen",
"preparation",
"and",
"treatment",
"staining",
"spectrum",
"analysis",
"techniques",
"white",
"blood",
"cells",
"animal",
"cells",
"proteins",
"medical",
"microbiology",
"hiv",
"microbial",
"pathogens",
"t",
"cells",
"hiv-1",
"proteasomes",
"spectrophotometry",
"biochemistry",
"cytophotometry",
"cell",
"staining",
"protein",
"complexes",
"cell",
"biology",
"viral",
"persistence",
"and",
"latency",
"virology",
"viral",
"pathogens",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"lentivirus",
"organisms"
] |
2019
|
Reiterative Enrichment and Authentication of CRISPRi Targets (REACT) identifies the proteasome as a key contributor to HIV-1 latency
|
Toxoplasmic retinochoroiditis appears to be more severe in Brazil , where it is a leading cause of blindness , than in Europe , but direct comparisons are lacking . Evidence is accumulating that more virulent genotypes of Toxoplasma gondii predominate in South America . We compared prospective cohorts of children with congenital toxoplasmosis identified by universal neonatal screening in Brazil and neonatal or prenatal screening in Europe between 1992 and 2003 , using the same protocol in both continents . Three hundred and eleven ( 311 ) children had congenital toxoplasmosis: 30 in Brazil and 281 in Europe , where 71 were identified by neonatal screening . Median follow up was 4 . 1 years in Europe and 3 . 7 years in Brazil . Relatively more children had retinochoroiditis during the first year in Brazil than in Europe ( 15/30; 50% versus 29/281; 10% ) and the risk of lesions by 4 years of age was much higher: the hazard ratio for Brazil versus Europe was 5 . 36 ( 95%CI: 3 . 17 , 9 . 08 ) . Children in Brazil had larger lesions , which were more likely to be multiple and to affect the posterior pole ( p<0 . 0001 ) . In Brazil , visual impairment ( <6/12 Snellen ) was predicted for most affected eyes ( 87% , 27/31 ) , but not in Europe ( 29%; 20/69 , p<0 . 0001 ) . The size of newly detected lesions decreased with age ( p = 0 . 0007 ) . T . gondii causes more severe ocular disease in congenitally infected children in Brazil compared with Europe . The marked differences in the frequency , size and multiplicity of retinochoroidal lesions may be due to infection with more virulent genotypes of the parasite that predominate in Brazil but are rarely found in Europe .
Toxoplasmic retinochoroiditis appears to be more common in Brazil than in Europe or North America and more severe . It is a leading cause of blindness in Brazil[1] but not in Europe or North America . [2] , [3] These differences are not adequately explained by high rates of postnatal or congenital infection in Brazil , as similar rates of infection have been observed in France and Eastern Europe . [4]–[8] Once infected , population-based studies of adolescents and adults , most of whom have postnatally acquired infection , report the risk of retinochoroiditis to vary from 2% in North Eastern Brazil to 25% in Southern Brazil . [4] , [9]–[11] No comparable studies have been done in Europe or North America but case series report risks of 0 . 3% to 1% in adults in the year or two after acquisition of infection . [12]–[14] The fact that retinochoroiditis results in more and larger lesions in South America than in Europe or North America is well accepted by clinicians but differences could reflect delayed presentation due to poor access to health care , referral bias , acquisition of infection from oocysts rather than tissue cysts , or exposure to a higher parasite load or to infection earlier in childhood in Brazil . [15] No published studies have directly compared ocular sequelae of toxoplasmic infection in Brazil with Europe or North America . Differences in the ocular sequelae of toxoplasmosis have been attributed to recent findings that distinctly different populations of Toxoplasma gondii exist in South America compared with Europe and North America . [16] , [17] Strains in Brazil appear to be more virulent , and have been identified in several patients with severe ocular disease . [18] , [19] In-vitro studies suggest that strain virulence affects triggering of the immune response , tissue penetration and the ability to encyst . As treatment is effective only during the tachyzoite phase , prior to encystment , this raises the possibility that the response to anti-toxoplasma treatment may differ between strains . [20] , [21] Possible clinical and policy implications of these findings could be the development of targeted treatment and preventive strategies depending on the prevailing parasite genotype . Our aim was to inform clinical practice and policy , by quantifying differences in ocular sequelae after toxoplasmic infection in Brazil compared with Europe . We directly compared cohorts of children with congenital toxoplasmosis identified by neonatal or prenatal screening . This approach minimized differences in the route and timing of infection as all fetuses were infected by transplacental transmission of tachyzoites . Health care access was also similar as screening was applied to all births , and we prospectively followed up children using the same protocol in both continents .
We prospectively recruited and followed up a cohort of live-born children with congenital toxoplasmosis identified between 1996 and 2003 by neonatal or delivery screening in 2 Brazilian centers , by neonatal screening in 3 European centers ( Sweden , Denmark , and Poland ) , and by prenatal screening in 10 European centers ( 7 in France , 2 in Italy , 1 in Austria ) . We also included children identified in a prospective national Danish neonatal screening study between 1992 and 1996 . [22] Details of screening methods and treatment in the European centers have been reported elsewhere . [22]–[25] In all centers , we excluded any mother or infant who was not first identified by universal screening and who could have been referred for testing because of symptomatology . The diagnosis of congenital toxoplasmosis was based on persistence of specific IgG antibodies after 11 . 5 months of age . [23] , [24] In Brazil , neonatal screening was based on testing of the neonatal Guthrie card blood spot using an IgM immunoassay ( VIDAS , Biomerieux ) in Campos Dos Goytacazes , and a commercial capture IgM fluorometric enzyme immunoassay ( Labsystems ) in Porto Alegre . For screening of mothers at delivery , an IgM immunoassay ( VIDAS , BioMérieux ) was used . [26] , [27] In Campos Dos Goytacazes , neonatal screening was offered to approximately 25% of the total births ( about 9 , 000 births per year ) who delivered on two days each week at public sector hospitals within the area . Study infants were recruited between 1999 and 2001 . In Porto Alegre , all patients were enrolled who had congenital toxoplasmosis identified between 1998 and 2003 by routine neonatal ( n = 17 ) or delivery ( n = 5 ) screening and who started follow-up in the Congenital Infections Clinic of Sao Lucas Hospital before 2 months of life ( 3 patients followed up elsewhere were excluded ) . Twelve of these 22 children were followed up in the public sector and 10 in the private sector . Postnatal treatment was prescribed for 12 months . Pyrimethamine ( 1 mg/kg/day ) and sulphadiazine ( 100 mg/kg/day ) given for the first six months were changed for the subsequent six months to alternating spiramycin and pyrimethamine-sulphonamide in Campos Dos Goytacazes and to a lower dose of pyrimethamine ( 1 mg/kg/3 times per week ) with sulphonamide ( 100 mg/kg/day ) in Porto Alegre . We used a standard questionnaire to record findings at routine ophthalmoscopic examinations before 4 months , at 12 months of age , and annually thereafter for all children in the cohort . Clinicians were asked to dilate the pupil and use indirect ophthalmoscopy . They used a standard proforma to record whether the retina was adequately visualized , describe the site of lesions using a diagram and text , and estimate lesion size in multiples of the optic disc diameter . We defined a recurrence as a new lesion that was detected for the first time more than one week after a previous adequate visualization of the retina . Analyses of lesion size were based on the size of the largest lesion at each new occurrence . Multiple lesions were based on the total number of separate lesions detectable at the last examination . We assumed that 2 included children ( in Brazil and Europe ) had retinochoroiditis although microphthalmia and severe vitreal opacities prevented adequate examination . One of us ( MRS ) categorized each eye according to whether visual impairment ( <6/12 Snellen ) was likely or not based on the last retinal diagram . MRS was blinded to visual acuity results . The incidence of first or recurrent lesions was calculated from the number of new detections and the child years of follow up using a generalized linear model with a Poisson distribution with a log link function and the 95% confidence interval was derived . Kaplan Meier curves were derived to describe the time to first detection of the first lesion after birth . If a child had no lesions , censoring occurred at the date of the last ophthalmic examination . The probability of lesions by 4 years of age was derived from the product-limit estimates of the survivor function . We used Cox proportional hazards regression to compare differences in the age at detection of the first retinochoroidal lesion in Brazil with European screening centers . As no significant difference was detected between the time to first lesion in European prenatal and neonatal screening centers we compared the size of lesions , risk of recurrence and multiplicity of lesions in the combined European cohort with children in Brazil . The mean size of the largest lesion at each new lesion occurrence was compared using a hierarchical linear model to take into account multiple occurrences in the same child . Results of autoregressive and compound symmetric models were examined to identify the best model ( i . e . that with the lowest AIC statistic ) . [28] Using the dataset combining all children , we had sufficient power to compare the timing of the first and recurrent lesions using methods for multiple failure time data . We used the PWP total time model with common effects[29] , [30] which assumes that a subject cannot be at risk for a kth eye lesion unless the ( k-1 ) th lesion has occurred . We chose this model for two reasons . First , the possibility that the infection process or immune response that affects time to initial eye lesion may also impact on times to subsequent lesions , and each period evaluated might present different baseline risks . Second , results were consistent with other approaches , but the PWP model yielded the most conservative hazards ratio . Research ethics approval was required for five cohorts where neonatal screening and follow up of positive results was not established routine practice: the two Brazilian centers , patients screened in Denmark between 1992 and 1995 , and in Sweden and Poland . In Brazil , screening and follow up was approved by the Ethical Review Board of the Pontificia Universidade Católica do Rio Grande do Sul , Porto Alegre , Brazil , the Ethical Review Board of Fundação Oswaldo Cruz ( FIOCRUZ ) in Rio de Janeiro , Brazil . Mothers were given information during pregnancy about the screening study and verbal consent was obtained for testing at the time of sampling . Refusal to participate in screen testing was recorded by removing the card used for toxoplasma testing . Separate written consent was obtained from parents of screen positive children to participate in the follow up study . None refused . In the Danish study from 1992–1995 , in Sweden and in Poland , mothers were given information during pregnancy or at delivery . At the time of blood sampling , verbal consent was obtained to screen testing and to follow up if found to be positive . Refusal to participate was recorded in writing on the filter paper card . This approach of using oral consent to testing and follow up was approved by the Scientific Ethics Committee of Copenhagen ( V100 . 1689/90 and L 02033 ) , the Ethical Committee at Huddinge Hospital , Stockholm ( No . 96-089 ) , and the Scientific Ethical Committee of the Karol Marcinkowski University of Medical Sciences in Poznań . In all other cohorts ( in France , Italy , Austria and Denmark from 1996 onwards ) screening and follow up of positive test results was part of universal , routinely provided care . Research ethics approval was not required at the time of the study in any of these centers because all three of the following criteria were met: the study did not involve any modification of routine practice and had no impact on patient care; data collected for the study were confined to information held in routine records of screen test results and follow up examinations; and data were anonymized prior to collation for the study . We specifically requested that the Research Ethics Committee of Great Ormond Street Hospital scrutinize the study and provide written confirmation that research ethics review was not required in the UK before the start of prospective data collection in 1996 .
The study compared 30 infected children identified by neonatal screening in Brazil with 71 children identified by neonatal screening ( 29 in Poland and 42 in Scandinavia ) and 210 by prenatal screening in Europe ( 171 in France , 15 in Italy and 24 in Austria ) . All had at least one ophthalmoscopic exam and reports of inadequate visualization of the retina were rare . No screen positive children were excluded in Brazil but in Europe , 5 children were excluded due to possible referral bias and a further 3 because they had no ophthalmoscopic exam . The shortest interval between a negative examination and a new lesion occurrence was 22 days . None of the mothers of children in the neonatal centers were treated during pregnancy , whereas 178/210 ( 85% ) in the prenatal centers were treated . All except 3 children ( all in Europe ) were treated postnatally , mostly for 1 year ( further details reported elsewhere ) . [22] , [23] , [27] The median age at the start of postnatal treatment was 3 days ( IQR: 0 , 15 ) in prenatal centers , 27 days ( IQR: 23 , 34 ) in European neonatal centers , and 47 days ( IQR 25 , 82 ) in Brazilian centers . The survival analyses in Figure 1 show that more children developed retinochoroiditis sooner in Brazil than in Europe . The hazard ratio shows a markedly increased risk of early retinochoroidal lesion in Brazil than in Europe ( Table 1; p<0 . 0001 ) . There was no evidence that the time to first lesion differed between European neonatal and prenatal screening centers ( p = 0 . 9601; Table 1 and Figure 1 ) . The hazard ratio for time to first retinochoroidal lesion in neonatal centers in Brazil compared with neonatal centers in Europe was 2 . 37 ( 95%CI: 1 . 62 , 3 . 47 ) . Retinochoroidal lesion recurred at an earlier age in Brazil than in Europe ( p = 0 . 0406; Table 1 ) . By 4 years of age , the probability of a second lesion among children with a first lesion was 43% in Brazil compared with 29% in Europe ( Table 1 ) . The risk of multiple recurrences was also greater in Brazil ( hazard ratio for the time from birth to multiple lesions: 3 . 44; 95%CI: 2 . 23 , 5 . 32 ) . There was no significant difference in the age at recurrence between European prenatal and neonatal screening centers ( p = 0 . 1740 ) ( Table 1 ) . Children in Brazil were more likely than those in Europe to have retinochoroidal lesions that affected the posterior pole and to have visual impairment predicted by the ophthalmologist assessing the retinal diagrams ( p<0 . 001; Table 1 ) . Relatively more children in Brazil than Europe had multiple lesions ( p<0 . 0001 ) . The size of the largest lesion was recorded for 62% ( 74/119 ) of first or recurrent lesions . Figures 2 A to C depict the size of the largest lesion for each lesion occurrence according to age at detection . The overall mean size was greater in Brazil than in Europe ( p<0 . 0001 , Table 1 ) . In both continents , lesion size decreased with the age at detection ( p = 0 . 0007 ) .
In Brazil , congenital toxoplasmosis resulted in more frequent and more severe ocular disease than in Europe . There is indirect evidence that these differences may be related to the predominance of virulent genotypes of the T . gondii parasite in Brazil . Randomized controlled trials are urgently needed in South America to determine treatment efficacy and the clinical effectiveness of neonatal screening .
|
Toxoplasma gondii is found throughout the world and is the most common parasitic infection in humans . Infection can cause inflammatory lesions at the back of the eye , which sometimes affect vision . These complications appear to be more common and more severe when people acquire infection in Brazil than in Europe or North America , but there have been no direct comparisons of patients identified and followed up in the same way . In this report , we compare children with congenital toxoplasmosis diagnosed at birth by universal screening in Europe and Brazil and followed up until the age of 4 . We found that Brazilian children had a 5 times higher risk than European children of developing eye lesions and their lesions were larger , more numerous and more likely to affect the part of the area of the retina responsible for central vision . Two-thirds of Brazilian children infected with congenital toxoplasmosis had eye lesions by 4 years of age compared with 1 in 6 in Europe . These stark differences are likely to be due to the predominance of more virulent genotypes of the parasite in Brazil , which are rarely found in Europe .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"public",
"health",
"and",
"epidemiology/infectious",
"diseases",
"infectious",
"diseases",
"pediatrics",
"and",
"child",
"health"
] |
2008
|
Ocular Sequelae of Congenital Toxoplasmosis in Brazil Compared with Europe
|
Tailored therapy aims to cure cancer patients effectively and safely , based on the complex interactions between patients' genomic features , disease pathology and drug metabolism . Thus , the continual increase in scientific literature drives the need for efficient methods of data mining to improve the extraction of useful information from texts based on patients' genomic features . An important application of text mining to tailored therapy in cancer encompasses the use of mutations and cancer fusion genes as moieties that change patients' cellular networks to develop cancer , and also affect drug metabolism . Fusion proteins , which are derived from the slippage of two parental genes , are produced in cancer by chromosomal aberrations and trans-splicing . Given that the two parental proteins for predicted fusion proteins are known , we used our previously developed method for identifying chimeric protein–protein interactions ( ChiPPIs ) associated with the fusion proteins . Here , we present a validation approach that receives fusion proteins of interest , predicts their cellular network alterations by ChiPPI and validates them by our new method , ProtFus , using an online literature search . This process resulted in a set of 358 fusion proteins and their corresponding protein interactions , as a training set for a Naïve Bayes classifier , to identify predicted fusion proteins that have reliable evidence in the literature and that were confirmed experimentally . Next , for a test group of 1817 fusion proteins , we were able to identify from the literature 2908 PPIs in total , across 18 cancer types . The described method , ProtFus , can be used for screening the literature to identify unique cases of fusion proteins and their PPIs , as means of studying alterations of protein networks in cancers . Availability: http://protfus . md . biu . ac . il/
Fusion proteins resulting from chromosomal translocations have important roles in several types of cancer and are extensively discussed in cancer research literature . The current biomedical literature resources , such as PubMed , comprise more than 28 million citations , with approximately 14 , 000 cancer-related papers from 2018 alone , and more than 3 million abstracts in total that mention ‘cancer’ . Similarly , the number of PubMed articles that mention ‘fusion proteins’ is also increasing rapidly . Thus , there is a growing need to catalog as well as curate this information . Hence , text mining-based methods to identify fusion proteins from PubMed are extremely important . Moreover , information regarding fusion proteins mentioned in the current literature have no standard format . The upshot is that identifying a certain fusion protein is non-trivial; for example , a fusion protein such as BCR–ABL1 is represented in variable forms in different texts [1–2] . These variations include the formatting of the fusion events themselves ( e . g . , BCR-ABL1 vs . BCR:ABL1 vs . BCR/ABL1 ) , and the keywords used to describe them ( fusions vs . fusion proteins vs . chimeric proteins vs . chimeras ) [3] . Moreover , when extracting protein–protein interactions ( PPIs ) of fusions , their actions can be described in varying ways ( e . g . , activate vs . interact vs . express vs . induce [4] ) . To collect information about multiple fusion proteins , we developed an in-house database , ChiTaRS [1] , which covers more than 11 , 000 cancer breakpoints . We continually mine the new literature for mentions of fusion proteins , their parent proteins and their associated PPI networks , so as to provide a constantly updated fusion protein database tool for the scientific community worldwide . Text mining is used in biology to reveal associations between genes and proteins , as described in the literature . Several earlier studies focused on developing text-mining approaches for modern medical text , in general , and cancer research , specifically . For example , several annotated corpora have been created to distinguish mutations , cancer processes , tumor suppressors , oncogenes and transcription factors [5–7] . Natural language processing ( NLP ) methods use named-entities that have different sequences/phrases of nouns and adjectives , while named-entities involved in relationships are designated by verbs . Syntactic analysis may be defined as the process of analyzing by NLP the strings of symbols that conform to the rules of formal language grammar . Further , several previous and currently available tools enable extracting a specific set of information from the literature . Examples of such biomedical text mining tools are: MetaMap [8] , WhatIzIt [9] , iHOP [10] , PubTator [11] and Gimli [12] . Moreover , to provide a constantly up-dated resource for research purposes , continuous evaluations and verifications are performed by the biomedical text mining community through multiple data assessment initiatives , e . g . BioCreative [13–14] , BioNLP [15] and i2b2 [16] , to name a few . Therefore , NLP and text mining of medical literature is a growing field that may provide a novel resource in fusion moieties and their cellular processes . Screening for fusion proteins and their PPIs is a relatively new field in biomedical text mining , and as such , only a limited number of relevant resources are available [17–18] . Some well-known databases of fusion proteins have been developed , such as ChiTaRS-3 . 1 [1] , ChimerDB 3 . 0 [19] , COSMIC [20] and TICdb [21] . However , no available biomedical resource can automatically extract the PPIs of fusion proteins that have been confirmed experimentally from scientific papers . Moreover , given that the two parental proteins for predicted fusion proteins were known , our previously developed method may be used to predict chimeric PPIs ( ChiPPIs ) associated with the fusion proteins [22] . Here , we present a text-mining approach , called Protein Fusions Server ( ProtFus ) that receives fusion proteins of interest , predicts their cellular network alterations by ChiPPI [22] and validates them by searching PubMed references . ProtFus mines the literature to identify unique cancer fusion proteins and their experimentally described PPIs in scientific publications . Thus , ProtFus provides unique ways for extracting and interpreting information present in public scientific resources . Our objectives and process for employing ProtFus were as follows: Thus , ProtFus can be used to validate from the biomedical literature , protein interactions of fusion proteins in cancer , which can then be empirically tested . Furthermore , the interactions can be used to validate the predicted ChiPPI networks [22] of multiple fusion proteins in different public databases .
The initial text validation was performed on input from PubMed to remove false positive results , followed by segregation into tokens . We performed stemming of words for sentences , followed by identifying named-entities within sentences with the ‘Porter2’ algorithm , using the ‘stemming’ package in Python [28] . The named-entities within sentences were blanked out to make them more generalized . This step was followed by using a bag-of-words representation [29] based on a frequency score ( FS ) for estimating the importance of selecting a token . For the bag-of-words representation , we used the FS threshold ( Ts , a ) ( Eq 1 ) : Ts , a=FSs , a×log10 ( τσ ) ( Eq 1 ) Here , FSs , a denotes the frequency of token s in article a , τ is the number of articles/abstracts , σ is the number of articles having s . This threshold is used to estimate the frequency score . We used the Naïve Bayes classification method to build the PPI extraction model [29] . This categorizes the tokens in abstracts and articles to either fusion proteins or interactions of fusion proteins , and assigns them to Medical Subject Headings ( MeSH ) terms . The ProtFus framework was developed on an Apache Dell R820 server , with 1TB RAM and with a back-end My-SQL database and 1PB of support data from ChiTaRS-3 . 1 [1] . The tool was developed using Python , whereas the interface was developed using CGI-Perl ( http://protfus . md . biu . ac . il/ ) . We used the N-gram model for detecting N-words at a time from a given sentence . An N-gram model is a model of "strings" or "sequences" in NLP by means of the statistical properties of N-grams , based on the appearance of letters , according to the Shannon information theory of likelihood [30] . Specifically , using a 2-gram method , all words in a sentence were broken down into two combinations , including unigrams and bigrams , i . e . , one- and two-word combinations [31] . For example , some possible sets of combinations were provided in Fig 2 . We extracted a set of bigrams , as well as combinations of 3- and 4-grams , from abstracts or full-text articles in order to train ProtFus to detect specific fusion protein instances . In addition , the instances of these tokens were counted in the back-end corpus [32] . A back-end text corpus was a structured set of texts that can be used for statistical analysis; it checks occurrences and validates linguistic rules in a specific context . In our case , the back-end corpus was used for performing background feature extraction using N-grams . Further , when FS was the standard feature score , a considerably high threshold ( Ts , a ) was given to tokens that appeared frequently in the corpus . Moreover , we also converted all abstracts or full-text articles into ‘similar-length’ feature vectors , where each feature represents Ts , a of the identified token . The rationale was that these feature vectors are further used for rescaling the overall feature score . Subsequently , we organized a bag-of-words representation of the feature vectors ( Table 3 ) . A bag-of-words was a representation of text that described the occurrence of words within a document . Its two components are a vocabulary of known words and a measure of the presence of known words . Thus , S1–S2 Tables ( Supporting information ) include the back-end corpus considered for tagging fusion proteins and their interactions . The word-token tagger had a back-end Synonyms ( with synonyms resource , S3 Table , Supporting information ) , whereas the RegEx tagger had a back-end Synonyms ( with rulebase , S4 Table , Supporting information ) . Likewise , Table 4 represents Precision and Recall for a retrieval step . The tokens were used to parse the texts for performing named-entity recognition ( NER ) [33] . NER locates and classifies named-entities in text into pre-defined categories . For example , the unannotated block of text ‘CRKL binds to BCR-ABL fusion protein’ can be annotated as ‘[CRKL] protein binds to [BCR] protein—[ABL] protein [fusion protein] key’ . This was followed by searching for a pattern like protein1-protein2 key or protein1/protein2 key or protein1:protein2 key ( e . g . , [BCR] protein—[ABL] protein [fusion protein] key ) . To associate a fusion event with a certain cancer we performed NER of ‘diseases’ . For example , in ‘BCR-ABL causes leukemia’ , we performed annotations such as ‘[BCR] protein-[ABL] protein [causes] action-verb [leukemia] cancer’ . The ProtFus method performs a search in PubMed abstracts or uploads a full text file that is based on a specific input text . For example , in the case of an input text , the result is displayed in a separate pop-up window , and the fusion proteins are highlighted . Similarly , in the case of PPIs among fusion proteins , the result window includes the input text , and the interactions are highlighted . Thus , another feature of ProtFus is direct searching using PubMed articles . Users can select from the drop-down menu of 100 , 200 or more , the number of articles to be considered for searching fusion proteins and their interactions . The result includes the abstracts of the articles that match best with fusion protein keywords ( e . g . , for BCR-ABL ) . This file can be further used for highlighting the fusion proteins and their interactions . Thus , Table 5 represents Precision and Recall for NER . We downloaded abstracts from PubMed to generate both training and test datasets . For a training set we used several datasets ( Table 1 ) . Other datasets served as test sets to evaluate the model that was built and a 10-fold cross-validation was performed , each time using 40% of the entities to train an extraction model and the remaining 60% to test it [33] ( see Supporting information ) . Tokenization is the task of chopping a character sequence and a defined document unit into pieces , called tokens , while perhaps throwing away certain characters , such as punctuation . Tokenization was performed using two specific taggers: The word-token tagger identified the property of words from the text for fusion proteins , like ‘fusion proteins’ , ‘fusion transcripts’ , ‘chimeric proteins’ , ‘chimeric genes’ and ‘fusion gene transcripts’; and "action words" for PPIs , like ‘activate’ , ‘block’ , ‘depend’ , ‘express’ and ‘interact’ . Similarly , the RegEx tagger recognizes and associates these word-tokens with their corresponding “literals” ( attributes ) . The tokenizer module segregates the text into ‘Biological’ , ‘Miscellaneous’ , ‘Function’ and ‘Literal’ tokens . For example , given the following text , “The small molecule BCR-ABL-selective kinase inhibitor imatinib is the single most effective medical therapy for the treatment of chronic myeloid leukemia” , the tokenization output is: Biological Tokens—‘small’ , ‘BCR-ABL-selective’ , ‘single’ , ‘medical’ and ‘chronic’; Miscellaneous Tokens—‘molecule’ , ‘kinase’ , ‘imatinib’ , ‘therapy’ , ‘treatment’ , ‘myeloid’ and ‘leukemia’; Function Tokens—‘effective’ and ‘inhibitor’; Literal Tokens—‘is’ , ‘the’ , ‘for’ and ‘of’ . Here , we present the structure of the corpus that was used for validation and testing . ProtFus considered all possible combinations of representing fusion proteins in text , by looking the back-end Rule-base as well as the fusion and PPI corpus . Now , we define the different keywords and tokens used by our method , as part of entity recognition . The back-end ‘Synonyms’ ( fusion corpus ) consists of ‘entity’ ‘relation token’ , such as ‘fusion’ ‘fusions , fusion transcript , fusion transcripts , fusion protein , fusion proteins , fusion gene , fusion genes’ , whereas ‘Synonyms’ ( PPI corpus ) consists of ‘entity’ ‘relation token’ , such as ‘Activate’ ‘activate , activates , activated , activating , activation , activator’ . Similarly , the back-end ‘Synonyms’ ( fusion ) consists of ‘Fusion proteins’ ‘Synonyms’ ‘Alternate representations’ , such as ‘EWS-FLI1’ ‘TMRPSS2-ERG’ ‘ews: fli1 , EWSR1: EWS , EWSR1/FLI1 , EWS/FLI-1’ . The ‘parser’ and the entity recognition module used ‘Rule-base’ and ‘Short Form Recognition’ back-end resources for identifying the final ‘best-suited’ entities and tokens , and also for filtering out the false positives . The ‘Rule-base’ ( for normalization ) consisted of ‘Rule’ ‘Input’ ‘Output’ ‘Reg Ex’ , such as ‘Normalization of case’ ‘BCR-ABL , bcr-abl , BCR:ABL , bcr:ABL , BCR/ABL , bcr/abl’ . Similarly , the ‘Rule-base’ ( for regular expression ) consisted of ‘Characteristics’ ‘Description’ ‘Rule’ ‘Reg Ex’ , such as ‘Fusion token’ ‘Tokens with fusion word occurrence’ should be separated by space/tokens ( ‘fusion|fusions|fusion genes|gene fusion|fusion protein|fusion transcripts’ ) etc .
We used a classical Naïve Bayes algorithm for training as well as extraction . The datasets were partitioned based on known fusion proteins and their interactors from the literature . This resulted in a training set ( 40% ) ( Table 1 ) and a set ( around 60% , when there was no reported fusion ) that was used for testing the algorithm in all PubMed references ( 2013–2017 ) ( Table 2 ) . There was no overlap between training and testing data . Subsequently , decisions were modeled for assigning labels to raw input data . This type of classification algorithm can also be thought of as a convex optimization problem , in which one needs to identify the minima of a convex function ρ , associated with an input vector v , having n entries ( Eq 2 ) , min ( ρ ( v ) ) v∈Zn ( Eq 2 ) Here , the objective function can be defined as Eq 3 , ρ ( v ) =Zn+1nσi=1nμ ( v;a ( i ) , b ( i ) ) ( Eq 3 ) where vectors a ( i ) ∈Zn are training instances ( 1≪n ) , y ( i ) ∈Zn that act as labels . To examine the accuracy of our algorithm , we performed a 10-fold cross-validation . For this purpose , we partitioned the input text into ten equal-sized sub-samples , of which five were retained for testing and five were used for model building . We also used the standard Precision , Recall and F-score values for validating the results . Precision ( P ) was defined as the fraction of retrieved instances that was relevant to the study . Precision can also be defined as the probability that randomly selected retrieved information is relevant ( Eq 4 ) . P=TPTP+FP ( Eq 4 ) Here , TP = true positive and FP = false positive . Similarly , Recall ( R ) is defined as the fraction of relevant instances that are retrieved for the study . Recall can also be defined as the fraction of the information relevant to the query that is successfully retrieved ( Eq 5 ) . R=TPTP+FN ( Eq 5 ) Here , FN = false negative . Finally , F-score is the harmonic means of precision and recall ( Eq ( 6 ) . F−score=2[P⋅RP+R] ( Eq 6 ) For example , if the standard query text contains 3 tokens that could be categorized as fusion proteins , and ProtFus identifies 2 of them , the accuracy can be calculated as: True ( standard ) tokens = n , y , n , a; Predicted ( by ProtFus ) tokens = n , n , n , a ( here , n = no token instance , y = token instance , a = noise ) . In this case , Precision = 0 . 75 , Recall = 0 . 75 and F-score = 0 . 75 . Similarly , the corresponding accuracy plot can be drawn by providing information about Precision , Recall , and F-score values , and the number of runs . ProtFus still had a high false-positive rate , due to the diverse corpus of texts and different forms of fusion mentions . However , this rate automatically decreased when the corpus was updated with better literals . To display the results of ProtFus in a user-friendly manner , we also built the Protein-Protein Interaction of Fusions ( PPI-Fus ) database ( http://protfus . md . biu . ac . il/bin/protfusdb . pl ) , supported by Apache Tomcat and My-SQL . This is an open source Big Data processing framework that supports ETL ( Extract , Transform and Load ) and machine learning , as well as graph generation . Some classical text mining tasks can also be performed by identifying biological , functional , literal and miscellaneous tokens , as well as chunks from text . Further , for the purpose of entity recognition , the word-token tagger has back-end Synonyms ( with a synonym resource ) , whereas the RegEx tagger has back-end synonyms ( with rule base ) . Further , in the case of identifying PPIs among fusion proteins , the pop-up result window included the input text with interactions highlighted . Another feature of ProtFus is direct searching using PubMed articles . Users can select from the given drop-down box , the number of articles to be considered for searching fusion proteins and their interactions . The result includes the abstracts of all the articles that best match with fusion protein keywords . This file can be further used for highlighting the fusion proteins and their interactions . Interestingly , the biological tokens correspond mainly to nouns; miscellaneous tokens may correspond to verbs , pro-verbs , adverbs . etc; function tokens correspond to verbs and adjectives; and literals correspond to conjunctions . Tables 4 and 5 represent the Precision and Recall for the retrieval step and NER , respectively ( see Methods ) . Similarly , Table 6 provides the overall accuracy of the Naïve Bayes classifier , whereas Table 7 represents a comparative analysis of the overall extraction rate of fusion proteins and their PPIs using ProtFus and a selection of other resources . This comparison showed that ProtFus performs much better in overall extraction , with 92% accuracy . Thus , the process of tokenization was a very important step in our script , as it filtered out essential tokens ( like protein and function tokens ) from non-essential ones ( like miscellaneous and literals ) for better data extraction . Considering discrete protein domains as binding sites for specific domains of interacting proteins , we catalogued the protein interaction networks of more than 11 , 000 cancer fusions in order to predict PPIs of fusion proteins using ChiPPI [22] . Mapping the effects of fusion proteins on cell metabolism and protein interaction networks reveals that chimeric PPI networks often lose tumor suppressor proteins and gain onco-proteins . As a case study , we compared the results generated by ProtFus with the interaction prediction accuracy of ChiPPI [22] . For example , in BCR-JAK2 fusion , ProtFus provided multiple hits regarding its occurrence in literature , such as , “It was demonstrated in preclinical studies that BCR-JAK2 induces STAT5 activation that elicits BCRxL gene expression” ( PMC3728137 ) , as correctly predicted by ChiPPI ( Fig 3 ) . To demonstrate the added value of the ProtFus tool , we performed a direct comparison with existing services . Table 7 represents the accuracy of ProtFus as compared to , ChimerDB-3 . 0 [19] , FusionCancer [23] and FusionDB [36] resources . ChimerDB-3 . 0 chooses fusion gene candidate sentences from PubMed , which are further used for training a machine learning model . FusionCancer and FusionDB do not use text mining for fusion prediction . However , we used these datasets for resource-based comparisons of predicted fusion proteins . For the set of 1817 fusion proteins that were tested , the efficiency of our algorithm was about 92% , including the false positive rate , with respect to extracting fusion proteins and their PPIs from text . We compared ProtFus with other tools , according to Precision , Recall and F-score . We also found the Receiver Operating Characteristic ( ROC ) curves useful for quantitative representation of our method . Fig 4 shows representative ROC curves generated in a typical experiment using ‘abstracts’ data . Compared to full-text articles , the extraction was better for abstracts . This is because the size of feature space is too large for full-text articles . For text classification purposes , abstracts may yield better results than full-text scientific articles . We also used various full-text journal corpus information for the purpose of evaluating our method’s performance over others [37] . Thus , text mining enables the inclusion of text-based data ( unstructured data ) in models that are subsequently for classification and clustering , and even anomaly detection . In our study , we used Bayesian learning to identify fusion proteins and their interactions . The effectiveness of ProtFus derives from the manner that it is used in specific cases . For example , the script can be updated to include different annotations associated with fusion proteins , which can be further used to study their properties .
This study focused on investigating large-scale biomedical text classification downloaded from PubMed . We utilized classical text-mining and machine learning strategies , and also Big Data infrastructure to design and develop a distributed and scalable framework . This was applied to identify fusion proteins and their interactions for classifying information extracted from tens of thousands of abstracts and full-text articles with associated MeSH terms . The accuracy of predicting a cancer type by Naïve Bayes using the abstracts was 92% . In contrast , its accuracy using the 103 , 908 abstracts ( for fusion proteins only ) , 90 , 639 full texts ( for fusion proteins only ) , 185 , 606 abstracts ( for fusion protein interactions ) and 353 , 535 full texts ( for fusion protein interactions ) was 88% . This study demonstrates the potential for text mining of large-scale scientific articles using a novel Big Data infrastructure , with real-time updating from articles published daily . ProtFus can be extended to other areas of biomedical research to improve searches in multiple medical records and medical data mining approaches .
|
Tailored therapy aims to cure cancer patients in a fully personal way . Thus , the continual increase in scientific information and , particularly , in published literature , drives the need for efficient methods of data mining to find unique personal genomic features and their connections . Fusion proteins , which are derived from the slippage of two parental genes or chromosomal translocations , are frequently drivers of cancers . We used our previously developed method for identifying chimeric protein–protein interactions ( ChiPPIs ) for multiple fusion proteins . In this paper , we present a validation approach , ProtFus , which receives fusion proteins of interest , predicts their cellular network alterations by ChiPPI and validates them by online literature searches . This process resulted in a set of 358 fusion proteins and their corresponding protein interactions , as training set . Next , for a test set of 1817 fusion proteins , we were able to identify 2908 previously published PPIs across 18 different cancer types . The described method can be used for screening the literature to identify unique cases of fusion proteins and their PPI networks , as means of studying alterations of protein networks for personalized approaches in cancers .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Conclusion"
] |
[
"cell",
"physiology",
"protein",
"interactions",
"protein",
"interaction",
"networks",
"data",
"mining",
"text",
"mining",
"artificial",
"intelligence",
"network",
"analysis",
"gene",
"types",
"information",
"technology",
"extraction",
"techniques",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"protein",
"extraction",
"proteins",
"fusion",
"genes",
"proteomics",
"biochemistry",
"cell",
"biology",
"genetics",
"biology",
"and",
"life",
"sciences",
"machine",
"learning",
"cell",
"fusion"
] |
2019
|
ProtFus: A Comprehensive Method Characterizing Protein-Protein Interactions of Fusion Proteins
|
Sensory and cognitive performance decline with age . Neural dysfunction caused by nerve death in senile dementia and neurodegenerative disease has been intensively studied; however , functional changes in neural circuits during the normal aging process are not well understood . Caspases are key regulators of cell death , a hallmark of age-related neurodegeneration . Using a genetic probe for caspase-3-like activity ( DEVDase activity ) , we have mapped age-dependent neuronal changes in the adult brain throughout the lifespan of Drosophila . Spatio-temporally restricted caspase activation was observed in the antennal lobe and ellipsoid body , brain structures required for olfaction and visual place memory , respectively . We also found that caspase was activated in an age-dependent manner in specific subsets of Drosophila olfactory receptor neurons ( ORNs ) , Or42b and Or92a neurons . These neurons are essential for mediating innate attraction to food-related odors . Furthermore , age-induced impairments of neural transmission and attraction behavior could be reversed by specific inhibition of caspase in these ORNs , indicating that caspase activation in Or42b and Or92a neurons is responsible for altering animal behavior during normal aging .
Neuronal dysfunction and cell death are hallmarks of age-related neurodegenerative disorders , such as Alzheimer's disease . Epidemiological and biomedical studies have demonstrated that both genetic and age-related factors are crucial for the development and progression of these disorders . Attempts to understand the underlying mechanism of functional alterations in neural circuits during “normal aging” are receiving considerable attention [1]–[3] and should provide new insights toward preventing and treating age-related disorders . However , our knowledge about whether and how neural circuits are remodeled and/or maintained during normal aging is still very limited . Caspases are highly conserved cysteine proteases , which function as central regulators of apoptosis [4] , [5] . Knockout mice lacking caspase-3 , caspase-9 , or the caspase activator , apaf-1 , all exhibit reduced neuronal apoptosis and brain malformation [6]–[11] , indicating that caspases are essential for normal brain development . In addition to their role in apoptosis , non-apoptotic roles for caspases , particularly in the nervous system , are being reported in vivo [12] . These roles include dendritic pruning in the developing Drosophila [13] , [14] , song habituation in birds [15] , [16] , synaptic long-term depression ( LTD ) in rat hippocampal neurons [17] , [18] , synaptic maturation of olfactory sensory neurons in mice [19] , and early synaptic dysfunction in a mouse model of Alzheimer's disease [20] , [21] . Although the essential role of caspases in developing and adult brains has been documented , the in vivo activation pattern of caspases has not yet been systematically investigated . In this report , we began with mapping caspase activity throughout the entire lifespan of the fruit fly . Using a genetic probe for caspase-3-like activity ( DEVDase activity ) [13] , we revealed spatiotemporal caspase activation in the adult brain . Moreover , we found that this caspase activation was particularly prominent in the antennal lobe ( AL ) and ellipsoid body , which are brain structures responsible for olfaction and visual place memory , respectively [22]–[24] . Interestingly , when we further investigated caspase activity in the antennal lobe , we determined that caspases were activated in an age-dependent manner in select ORNs , particularly in Or42b and Or92a neurons that are essential for mediating innate attraction to food odors [25] , and that elevation of caspase activity caused ORN death . Furthermore , two-photon calcium imaging of projecting neural dendrites ( secondary neurons receiving input from ORNs ) indicated that aging reduced sensitivity of the related olfactory glomeruli , which could be suppressed by the expression of p35 , a caspase inhibitor . Lastly , we found that the age-related impairment of innate attraction behavior was also significantly suppressed by the inhibition of DEVDase in Or42b and Or92a neurons . Taken together , our data suggest that caspase activation in the aging brain is spatio-temporally regulated and actively contributes to age-related alterations of neural function .
To monitor DEVDase activity in the brain of the adult Drosophila , we used a genetically encoded DEVDase probe consisting of a transmembrane mouse CD8 ( mCD8 ) protein and a yellow fluorescent protein ( Venus ) linked by the caspase-3-cleavage sequence derived from human poly ADP ribose polymerase ( PARP ) [13] ( Figure 1A ) . The activated form of DEVDase cleaves this probe , known as mCD8::PARP::Venus , into two fragments . Moreover , an antibody against cleaved PARP ( anti-cPARP Ab ) can specifically detect one of these two fragments; the immunohistochemical cPARP signal thus generated reflects levels of activated DEVDase . We expressed mCD8::PARP::Venus in postmitotic adult neurons marked by elav-Gal4 and found that the brains of very young ( 0–1 day old ) and very old ( 30–45 days old ) flies exhibited higher cPARP signaling frequency than other age groups ( Figure 1B ) . In young adult brains , cPARP signals were primarily detected in the subesophageal ganglion ( SOG ) and in the midline region; however , the intensity of these signals varied in the SOG of individual brains ( Figure 1C and 1D ) . The cPARP brain pattern was similar between males and females , although cPARP appeared more frequently in males than in females ( Figure 1B , 1F and 2C ) . These results are consistent with previous findings obtained using the terminal deoxynucleotidyl transferase dUTP nick end labeling ( TUNEL ) assay and an antibody aimed to detect active forms of caspase-3 [26] . In contrast , brains of aged flies tended to exhibit cPARP in the dorso-medial corner of the AL and in the ellipsoid body ( Figure 2A–2C ) . Other neuronal processes showed cPARP signals in the aged brain , but the labeling appeared to be random ( Figure 2D and 2E; some data not shown ) . Importantly , we determined that the cPARP pattern in the AL was highly stereotyped in aged flies of both sexes ( 32 . 2% of male brains and 8 . 2% of female brains , at 45 days post-eclosion ) ; hence , we focused on the AL neural circuit . The AL is the first olfactory center in the Drosophila brain that consists of ∼50 glomeruli , which are ball-shaped synaptic structures that receive axons of ORNs and dendrites of projection neurons ( PNs ) and are interconnected by local interneurons ( LNs ) [22] . To identify the neuronal subtypes with DEVDase activity , we expressed mCD8::PARP::Venus using pebbled-Gal4 ( all ORNs ) , NP1227-Gal4 ( GABAergic LNs ) , and GH146-Gal4 ( two-thirds of the PNs ) . Only pebbled-Gal4 generated reproducible cPARP signals in aged fly brains , which were suppressed by p35 ( Figure 3 ) . Further , we used 17 Or-Gal4 drivers to express mCD8::PARP::Venus in each ORN subtype . Surprisingly , we found that DEVDase was frequently activated in the axons of the Or42b , Or92a , and Or35a neurons but rarely in the other classes of ORNs that we tested ( Figure 4A–4C ) . Or47b neurons also showed DEVDase activation , but cPARP signal intensity in these cells was very low . These data indicate that DEVDase activation is age-dependent in specific subsets of ORNs . Strong activation of DEVDase in a cell body typically leads to apoptosis [5] . To determine whether DEVDase activation in Or42b and Or92a neurons caused apoptosis , we examined the activation state of DEVDase in ORN cell bodies located in the third segment of the fly antenna . In aged flies , among mCD8::PARP::Venus-positive Or42b and Or92a neurons , we found that a small fraction of neurons were positive for cPARP ( Figure 5A ) : 6 . 7% of Or42b ( n = 60 cells ) and 3 . 1% of Or92a ( n = 65 cells ) . On the other hand , none of the ORN cell bodies showed cPARP signals in young flies ( 0% Or42b neurons , n = 71 cells; 0 of Or92a neurons , n = 81 cells ) . In addition , some Or42b neurons are positive for both TUNEL and cPARP signal ( Figure 5B ) . To further examine death of ORNs of aged flies , we expressed a nuclear-localized enhanced cyan fluorescent protein ( ECFP ) ( Histone H2B::ECFP ) in each type of ORN and found that there was a significant decrease in Or42b and Or92a neurons during aging ( Figure 5C and 5D ) , while the number of Or85a neurons remained constant ( Figure 5E ) . The expression of several apoptosis inhibitors ( p35; the dominant-negative form of Drosophila caspase-9 Dronc , Dronc-DN [27]; and microRNA for reaper , hid , and grim , miRHG [28] ) effectively reversed the trend of an age-dependent decrease in Or42b and Or92a neurons ( Figure 5C and 5D ) . In contrast , the expression of these apoptosis inhibitors did not affect neuron numbers in young flies ( Figure S1 ) . These data suggest that Or42b and Or92a neurons die , at least in part , by caspase-mediated cell death in an age-dependent manner . Since the above-mentioned ORN subtypes die with DEVDase activation during aging , we hypothesized that odor-evoked behavior through the Or42b and Or92a neurons would be impaired with age . To test this possibility , we measured innate attraction behavior to apple cider vinegar in young and aged flies . Apple cider vinegar excites six glomeruli including DM1 and VA2 , which are innervated by the axons of Or42b and Or92a , respectively [25] . We found that attraction to apple cider vinegar was significantly decreased in aged flies , and that this effect was reversed by p35 expression in Or42b and Or92a neurons ( Figure 6A ) . These results clearly indicate that DEVDase activation in Or42b and Or92a neurons is the main cause of age-related impairments to innate attraction behavior . Lastly , we wanted to investigate age-dependent change of glomerular sensitivity by using two-photon microscopy imaging with a genetically encoded calcium sensor , GCaMP [29] . We measure the sensitivity of a given glomerulus by monitoring its output projection neurons . Specifically , we image the dendritic calcium levels of PNs innervating each glomerulus , we applied apple cider vinegar to the flies bearing GH146-LexA and LexAop-GCaMP1 . 3-ires-GCaMP1 . 3 . We found that , in young fly brains , the DM1 glomerulus was robustly activated in response to apple cider vinegar in a concentration-dependent manner ( Figure 6B and 6C ) . In contrast , the DM1 glomerulus was only weakly activated even at high concentrations of apple cider vinegar in aged flies . Moreover , the expression of p35 in Or42b neurons increased sensitivity of DM1 to vinegar ( Figure 6D and 6E ) . These results are consistent with our observations on attraction behavior ( Figure 6A ) , and indicate that the olfactory response of the DM1 glomerulus is impaired during aging due to DEVDase activation in Or42b neurons .
In the current study , we demonstrate that normal aging increases caspase activity , leading to age-related cell death , reduced olfactory sensitivity , and impaired innate attraction behavior . Since caspase-3 activity appears to contribute to synaptic LTD in the rat hippocampus [17] , [18] and to early synaptic dysfunction in mouse models of Alzheimer's disease [20] , further studies of age-related increases in caspase activity and its role in Drosophila neuronal excitability and cell death are warranted . Our discovery that specific types of ORNs show DEVDase activation and cell death is the first example of age-related , stereotyped cell death of neurons within a specific network . The activation of the DM1 and VA2 glomeruli , which are innervated by Or42b and Or92a neurons , respectively , is essential for innate food attraction behavior [25] . Thus , our observations may help to explain age-related changes in innate animal behavior . In addition to the olfactory system , we found stereotyped caspase activation in the ellipsoid body , the brain region involved in olfactory memory consolidation [23] and visual place memory [24] . This observation could imply the possible contribution of caspase activation to age-related memory impairment ( AMI ) . Like other animals , aged flies exhibit AMI , which corresponds to an increase of cAMP-dependent protein kinase ( PKA ) in the mushroom body but not in the ellipsoid body [30] . Because caspase is required in synaptic LTD [17] , it might be interesting to investigate the possible implication of caspase activation in the ellipsoid body for olfactory or visual memory and whether its role is apoptotic or non-apoptotic . The results of our current study reveal an interesting phenomenon in that age-related caspase activation only occurred in select ORNs . One possible explanation for this is the continuous activation of Or42b and Or92a neurons by food odors . As previously discussed , Or42b and Or92a respond to odors that flies recognize as food , such as apple cider vinegar [25] . Under regular experimental conditions , flies are cultured in a food-containing vial leading to continuous activation of Or42b and Or92a for the duration of a fly's lifespan . To test whether this continuous ORN activation was responsible for the eventual age-related caspase activation in these neurons , we aged flies in a vial containing yeast paste and examined cPARP signal in Or42b neurons . Interestingly , we found that the age of onset and strength of caspase activation in Or42b neurons was similar to what we found in flies that had been cultured in normal food ( data not shown ) , suggesting that a continuous food odor is not solely responsible for inducing age-related caspase activation . As for the involvement of neural activity in the maintenance of ORN axons [31] , further studies of culturing conditions containing restricted odors and the genetic manipulation of ORN neural activity would help to clarify these issues . In addition to neuronal activity , aging itself might produce ORNs with differential sensitivity to neuronal excitability or toxicity . This idea is supported by a report from Tonoki et al . ( 2011 ) , showing that forced expression of a truncated form of human Machado-Joseph disease protein with an expanded polyglutamine domain in the adult Drosophila eye at 20–24 days after eclosion causes more severe neurodegeneration than expression at 0–4 days of age [32] . These observations suggest that neuronal identity , including sensitivity to a toxic factor generated by age-related neuronal excitability , may be continuously changing over the course of a fly's lifecycle . Differential expression of the effector caspases , drICE and Dcp-1 , may determine the spatiotemporal specificity of caspase activation . Previous studies have suggested that expression levels of these caspases reflect the apoptotic potential of cells , and that drICE is more effective than Dcp-1 to induce apoptosis [33] . Interestingly , it has been shown that activation of drICE and Dcp-1 can only be detected in degenerating dendrites but not in the cell body of Drosophila C4da neurons [34] . We also previously reported a similar phenomenon in mice where caspase-3 could be detected in the developing axons of olfactory sensory neurons but not in the cell body [19] . In the current study , we found that caspases were activated in both the axon and cell body of a subset of Or42b neurons that eventually go on to die in an apoptotic manner , while the subset that did not show elevated levels of caspases went on to survive . Therefore , we expect that both drICE and Dcp-1 were likely activated in dying Or42b neurons , while either drICE or Dcp-1 was activated in the degenerating axons of surviving Or42b neurons . It has been reported that the Or42b and Or92a genes are the most conserved in the drosophilid olfactory subgenome and seem to be utilized to detect odors from wild lily ( Solomon's lily ) in other drosophilid species [35] . Thus , Or42b and Or92a seem to possess the most fundamental function among the ∼50 types of olfactory neurons . Moreover , this suggests that caspase activation in these neurons might have a greater impact on animal behavior . Therefore , we believe that this would be an ideal experimental paradigm to investigate age-dependent changes of innate behaviors . Lastly , our current findings suggest that while it is clear that caspase activation plays a crucial apoptotic role in the adult olfactory circuit , caspase activation may also have non-apoptotic functions . This is in light of that fact that while we were only able to detect a few TUNEL-positive cells among cPARP-positive ORNs , we noted a significant reduction in odor-evoked neural activity in the DM1 glomerulus of aged flies . Richard et al . recently identified an age-dependent disruption of a specific synaptic layer in the mouse olfactory bulb without any detectable neuronal loss [36] . In addition , it has been shown that caspase-9 is activated in aged olfactory bulb neurons without affecting the number of these cells [37] . These observations , together with our study , prompt questions concerning the ecological and pathological significance of caspase activation , or synaptic dysfunction , in specific groups of neurons or synapses of the adult olfactory circuit . Investigating the relationship between age-related alterations in neural circuits may provide clues to understanding the neural basis of impaired sensory and cognitive performance during normal aging and senile dementia .
The following transgenic lines were used: elav-Gal4 , Or-Gal4 , ( Bloomington Stock Center ) , NP1227-Gal4 ( Kyoto Drosophila Stock Center ) , UAS-mCD8::PARP::Venus [13] , pebbled-Gal4 [38] , GH146-Gal4 [39] , UAS-miRHG [28] , UAS-reaper [40] , UAS-p35 ( a gift from Bruce Hay ) , UAS-Dronc-DN [27] , UAS-mCD8::GFP [41] , UAS-H2B::ECFP [42] , LexAop-GCaMP1 . 3-ires-GCaMP1 . 3 [25] , and GH146-LexA [43] . All flies were maintained in a 25°C incubator and transferred to vials with fresh food every 3 to 4 days . Immunohistochemistry of the Drosophila adult brain was performed as previously described [44] . To stain ORN cell bodies , we dissected antennae from flies , fixed them in 4% ( vol/vol ) paraformaldehyde/0 . 3% ( vol/vol ) phosphate-buffered saline with Triton X-100 ( PBT ) at room temperature ( R . T . ) for 30 min , mounted them in OCT , and cut 14 µm-thick sections on a cryostat . Slides were then re-fixed with 4% ( vol/vol ) paraformaldehyde/0 . 3% ( vol/vol ) PBT at R . T . for 30 min , washed with 0 . 3% ( vol/vol ) PBT , and labeled using standard techniques . Antibodies used include rat anti-mouse CD8 antibody ( 1∶100 , MCD0800 , Invitrogen ) , rabbit anti-cleaved PARP ( Asp214 ) antibody ( 1∶100 , #9541 , Lot . 7 , Cell Signaling ) , rabbit anti-cleaved PARP antibody [Y34] ( 1∶100 , ab32561 , Abcam ) , nc82 mouse monoclonal antibody ( 1∶40 , Developmental Studies Hybridoma Bank ) , anti-rat Alexa488 ( 1∶250 ) , anti-rabbit Cy3 ( 1∶1000 ) , and anti-mouse Cy5 ( 1∶1000 ) ( Jackson Laboratory ) . Confocal images were captured using a Leica SP5 confocal microscope . TUNEL assay of antennal cryosections was performed using an In Situ Cell Death Detection Kit , TMR red ( Roche ) . Tissues were mounted in SlowFade Gold Antifade Reagent with DAPI ( Life Technologies ) . Cantonized w1118 [w ( CS10 ) ] flies were used as behavioral controls in our experiments . The flies used for behavioral assays were out-crossed to the w ( CS10 ) background . All fly stocks were maintained at 25°C and 70% relative humidity under a 12/12 h light-dark cycle . For behavioral studies , about 50 male flies were placed into food vials and transferred to fresh food vials every 3 or 4 days until the age for behavioral assay was reached . Behavioral assays were performed under dim red light at 25°C and 70% relative humidity . Attraction to apple cider vinegar was measured for 3 or 30-day-old flies . Briefly , flies of each type were loaded into a T maze in which they could make a choice between two arms . Flies were allowed 2 min to choose between an odor of apple cider vinegar or air . Apple cider vinegar was diluted in water to 0 . 3% ( v/v ) , which elicits robust attraction behavior in 3-day-old flies . The performance index ( P . I . ) was defined as the ratio of the difference in number of flies that chose the air laced with or without apple cider vinegar odor to the total number of flies that chose either side . Calcium imaging was performed as described [25] , [29] , [45] . For odor stimulation experiments , a constant airflow of 1 L/min was applied to the antennae via a tube of 12 mm diameter . Odor onset was controlled by mixing a defined percentage of carrier air redirected through odor bottles ( presented as percent saturated vapor pressure , or %SV ) as previously described [45] . Third segments of the antennae were dissected in phosphate-buffered saline ( PBS ) and mounted with FocusClear mounting solution ( Cedarlane Laboratories ) . All cell images ( H2B::ECFP ) were taken live by a Leica SP5 confocal microscope within 15 min . Cell numbers were manually counted with ImageJ software and statistical analyses were performed using Stastical Package for the Social Sciences ( SPSS 16 . 0 ) ( IBM ) and Prism software ( GraphPad ) .
|
The approaching era of an “aging society” is receiving considerable attention amongst biomedical researchers in advanced nations . In order to understand the molecular mechanisms underlying age-related alterations of neural circuitry , we focused on caspase-3 , a cysteine protease that induces apoptotic cell death , using the fruit fly Drosophila melanogaster , a model often used to study aging due to a short lifespan of approximately 30–60 days . Here , we describe the spatiotemporal activation of caspase-3 in aged fly brains and show that caspase-3 is specifically activated in select olfactory neurons essential for innate odor attraction behavior . Furthermore , we discuss how inhibition of caspase-3 activation in those select olfactory neurons can rejuvenate the sensitivity of innate attraction behavior in aged flies . These findings suggest that caspase-3 plays an active role in producing age-related alterations to neuronal physiology and circuit function associated with animal behavior .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"arthropoda",
"animal",
"models",
"cell",
"death",
"invertebrates",
"drosophila",
"melanogaster",
"model",
"organisms",
"cell",
"biology",
"olfactory",
"system",
"biology",
"and",
"life",
"sciences",
"cell",
"processes",
"sensory",
"systems",
"molecular",
"cell",
"biology",
"drosophila",
"neuroscience",
"animals",
"insects",
"organisms",
"research",
"and",
"analysis",
"methods"
] |
2014
|
Caspase Inhibition in Select Olfactory Neurons Restores Innate Attraction Behavior in Aged Drosophila
|
CC2D1A and CC2D1B belong to the evolutionary conserved Lgd protein family with members in all multi-cellular animals . Several functions such as centrosomal cleavage , involvement in signalling pathways , immune response and synapse maturation have been described for CC2D1A . Moreover , the Drosophila melanogaster ortholog Lgd was shown to be involved in the endosomal trafficking of the Notch receptor and other transmembrane receptors and physically interacts with the ESCRT-III component Shrub/CHMP4 . To determine if this function is conserved in mammals we generated and characterized Cc2d1a and Cc2d1b conditional knockout mice . While Cc2d1b deficient mice displayed no obvious phenotype , we found that Cc2d1a deficient mice as well as conditional mutants that lack CC2D1A only in the nervous system die shortly after birth due to respiratory distress . This finding confirms the suspicion that the breathing defect is caused by the central nervous system . However , an involvement in centrosomal function could not be confirmed in Cc2d1a deficient MEF cells . To analyse an influence on Notch signalling , we generated intestine specific Cc2d1a mutant mice . These mice did not display any alterations in goblet cell number , proliferating cell number or expression of the Notch reporter Hes1-emGFP , suggesting that CC2D1A is not required for Notch signalling . However , our EM analysis revealed that the average size of endosomes of Cc2d1a mutant cells , but not Cc2d1b mutant cells , is increased , indicating a defect in endosomal morphogenesis . We could show that CC2D1A and its interaction partner CHMP4B are localised on endosomes in MEF cells , when the activity of the endosomal protein VPS4 is reduced . This indicates that CC2D1A cycles between the cytosol and the endosomal membrane . Additionally , in rescue experiments in D . melanogaster , CC2D1A and CC2D1B were able to functionally replace Lgd . Altogether our data suggest a functional conservation of the Lgd protein family in the ESCRT-III mediated process in metazoans .
CC2D1A ( Coiled-coil and C2 domain-containing protein 1A ) /LGD2 and CC2D1B/LGD1 belong to the evolutionary conserved Lgd protein family that is present in the genomes of all metazoans [1] . Members of this family contain four tandem repeats of the DM14 domain and one C2 domain . The D . melanogaster ortholog Lgd ( Lethal ( 2 ) giant discs ) was shown to be involved in the trafficking of the Notch receptor and other transmembrane proteins through the endocytic pathway . Loss of its function results in an ectopic and ligand-independent activation of the Notch pathway in several tissues that leads to over-proliferation of imaginal disc cells [1–4] . Therefore lgd was classified as a hyperplastic tumour suppressor gene [5] . Moreover , its loss causes enhanced activation of the D . melanogaster BMP signalling pathway ( Dpp pathway ) during oogenesis [6] . Several rather diverse functions have been described for CC2D1A . It was first identified in a large-scale screen to identify genes that activate the NFκB pathway in HEK293 cells [7] . Later , the function in the canonical IKK pathway was confirmed [8] . Additionally , CC2D1A appears to act as a transcriptional repressor of the dopamine receptor gene DRD2 [9] and the serotonin receptor gene 5-HT1A [10] . Likewise , CC2D1B functions as a repressor for the 5-HT1A gene [11] . Moreover , CC2D1A seems to be involved in the regulation of signalling pathways . During EGFR signalling it acts as a scaffold protein to recruit and activate PDK1/Akt [12] . In line with this observation is that silencing of CC2D1A inhibits growth of EGFR induced lung cancer cells [13] . It is a positive regulator of the cAMP/PKA pathway , where it is required for PKA activation and regulation of PDE4D [14 , 15] . In innate immunity , CC2D1A modulates the TLR3 and TLR4 signalling pathways [16] and the RLR pathways [17] . Centrosome associated CC2D1A appears to regulate centriole cohesion by preventing premature cleavage in HeLa cells [18] . However , most of these ascribed diverse functions are based on cell culture experiments and it is not known whether the CC2D1 proteins perform them in vivo . Several mutations in CC2D1A are linked to severe forms of intellectual disability and autism spectrum disorder in humans [19 , 20] . Cc2d1a deficiency in mice leads to an even severer phenotype [15 , 17 , 21] . Mutant mice die postnatally within a few minutes [21] , hours [15] or within a day after birth [17] due to difficulties in breathing . The underlying cause of the defect was suspected to be a defect in maturation of synapses [21] or dysregulation of the cAMP/PKA pathway [15] . However , the analysis was performed with conventional knockout mice and it remains unclear which mutant tissue contributes to the breathing defect . In contrast to CC2D1A , virtually nothing is known about the function of CC2D1B in vivo . In D . melanogaster , Lgd physically interacts with Shrub ( CHMP4 in mammals ) , the major subunit of the ESCRT-III complex , which is required for the formation of intraluminal vesicles ( ILVs ) during endosomal maturation [22] . ESCRT-III together with four other ESCRT-complexes generates the ILVs in maturing endosomes ( MEs ) , which—as a result—become multivesicular bodies ( MVBs ) . In D . melanogaster , ILV formation is required for the termination of signalling through several signalling pathways , among them the BMP and EGFR-pathways . Moreover , it prevents the ectopic activation of the Notch pathway [23–25] . Whether this is also the case in mammals is not known . The other ESCRT-complexes are termed ESCRT-0 , I , II , and the VPS4 complex . ESCRT-0—II act in sequence to concentrate ubiquitinated cargo and to assemble ESCRT-III at the sites of cargo concentration . ESCRT-III consists of four subunits ( CHMP6 , CHMP4 , CHMP3 and CHMP2 ) as well as several auxiliary factors ( reviewed in [26] ) . Three CHMP4 genes are present in humans , termed CHMP4A , B and C . Mice lack CHMP4A [27] . CHMP6 is activated by ESCRT-II and induces the polymerisation of CHMP4s into a filament that is capped by CHMP3 and CHMP2 . The filament is disassembled by the AAA ATPase VPS4 into monomers during vesicle abscission . CHMP4 monomers exist in a “closed” inactive form in the cytosol and can be recruited to the limiting membrane for another round of polymerisation . Thus , CHMP4s cycle between a monomeric closed form in the cytosol and an “open” polymerised form at the endosomal membrane . The cycling is the reason for the finding that the majority of CHMP4s are located in the cytosol upon antibody staining . Only if the function of Vps4p is abolished , ESCRT-III proteins strongly accumulate at the endosome and can be detected there in antibody staining [28] . Although clearly involved in endosomal trafficking in D . melanogaster , no definite link to endocytic trafficking has been established for the mammalian orthologs of Lgd , CC2D1A and CC2D1B . Rescue experiments in D . melanogaster suggested functional similarities , since murine CC2D1A and CC2D1B were able to replace Lgd during wing development [1] . However , it turned out later that the rescue assay was not suitable to test the ability of the Lgd orthologs to functionally replace Lgd [22] . Another assays suggested that down-regulation of CC2D1A in HeLa cells weakly decreased EGF and TF endocytosis [29] . A yeast two-hybrid screen revealed an interaction between CC2D1A and CHMP4 proteins [30] . This interaction was further refined by Martinelli et al . by showing that CC2D1A might regulate CHMP4B polymerisation [31] . The same group could also show that the over-expression of CC2D1A inhibits CHMP4B dependent HIV-1-budding in cells . Correspondingly , siRNA mediated depletion of CC2D1A increased HIV-1-budding under certain conditions [32] . Taken together , these experiments suggest that mammalian members of the Lgd gene family might also participate in the regulation of the endocytic pathway . Like in D . melanogaster , the Notch signalling pathway plays a pivotal role during development and tissue homeostasis in mammals ( reviewed in [33] ) . In the gut epithelium , the role of the Notch pathway is particularly well understood . It is required for the maintenance of proliferation of intestinal stem cells and for the differentiation of the progenitor cells in the crypt . Inhibition of Notch signalling results in the arrest of crypt cell proliferation and guides all crypt cells into a goblet cell fate [34] . Conversely , ectopic Notch signalling in the embryonic and adult intestine leads to ectopic proliferation of crypt progenitor cells and impairs goblet and enteroendocrine cell differentiation [35 , 36] . The Notch signalling pathway is activated through direct binding of its ligands ( reviewed in [37] ) . The binding elicits the separation of the intracellular domain of the receptor ( NICD ) through two proteolytic cleavages by ADAM10 and the γ-secretase complex , which subsequently travels to the nucleus and associates with the CSL transcription factor to activate expression of target genes , such as Hes1 . Here , we analyse the role of CC2D1A and CC2D1B in mouse with the focus on the questions whether CC2D1 proteins are involved in the endocytic pathway and the regulation of the activity of Notch signalling as observed for Lgd in D . melanogaster . For this purpose , we have generated conditional knockout mouse models for both genes and analysed their function . We show for the first time that the postnatal death of Cc2d1a deficient mice is caused by tissue-specific absence of the gene function in the nervous system . In contrast , Cc2d1b does not appear to be essential for embryogenesis or survival or fertility of mice . Furthermore , we find that Notch signalling is not altered in the intestine of Cc2d1a or Cc2d1b deficient mice compared to wild type . Murine embryonic fibroblasts ( MEFs ) derived from mutant animals do not show detectable alterations in the endocytic trafficking and degradation of the Notch receptor . We observe that CC2D1A , like D . melanogaster Lgd , locates in the cytosol , but accumulates at MEs when VPS4 function is reduced . This finding , together with cell fractionation data , indicates that CC2D1 proteins are temporally associated with the ME . EM analysis of MEFs reveals an increase of the size of the endosome in Cc2d1a deficient cells , indicating a defect in endosomal morphogenesis . In D . melanogaster , expression of CC2D1A and CC2D1B under the control of the endogenous lgd promoter can rescue the loss of function of lgd . Altogether , the results provide strong evidence that the endosomal function of the Lgd protein is conserved in mammals .
To characterise the function of CC2D1B in vivo we generated a Cc2d1b conditional null allele by breeding the KOMP mouse strain C57BL/6N-Cc2d1btm1a ( KOMP ) Wtsi with a Flp deleter strain to generate a Cc2d1b allele where exon 3 is flanked by loxP sites ( Fig 1A ) . Cre-mediated recombination deletes the floxed exon , leading to a frame-shift mutation and a premature stop codon after 102bp of missense sequence . The predicted truncated protein consists of 56 amino acids and contains no known domains . We confirmed Cc2d1b deficiency by genotyping PCR , RT-PCR and immunoblotting ( Fig 1B–1D ) . For immunoblotting , we generated a polyclonal antibody directed against the N-Terminus of CC2D1B ( aa 1–253 ) . The truncated protein was not detectable , suggesting low affinity of the polyclonal antibody to epitopes in the truncated protein , nonsense-mediated mRNA decay or instability of the truncated CC2D1B protein . Nevertheless , a putative dominant negative effect of such a fragment is highly unlikely as the fragment lacks all known functional domains . The genotype distribution of offspring from heterozygous breedings matched the expected Mendelian ratio ( n = 30; wild type = 6 , heterozygous = 17 , homozygous = 7; χ2 = 0 . 59 , p = 0 . 74 ) . Homozygous animals are viable and fertile with no obvious abnormalities and were kept as a homozygous line . In summary , loss of Cc2d1b function does not lead to an obvious phenotype under the used conditions . One possibility why the loss of Cc2d1b function has no detectable effect on viability of mice is that the expression of CC2D1A is up-regulated to compensate for its loss . We tested this possibility in Western blots and found that the expression of CCD1A was unaffected in Cc2d1b deficient MEFs , instead of elevated as would be expected ( S1A and S1C Fig ) . Conversely , the loss of Cc2d1a caused a reduction of the CC2D1B level to approximately 60% of the wild type level ( S1A and S1B Fig; see below for Cc2d1a mutant cells ) . Thus , the lack of a mutant phenotype of Cc2d1b mutants cannot be explained by a compensatory up-regulation of the expression of CC2D1A . To characterise the function of CC2D1A in vivo we generated a Cc2d1a conditional null allele where the exons 7 and 14 are flanked by loxP sites ( Fig 2A ) . Cre-mediated recombination leads to deletion of the floxed DNA segment . In the resulting Cc2d1a transcript exon 6 is spliced to exon 15 , leading to a frame-shift mutation and a premature stop codon after 120bp of missense sequence . The truncated protein consists of 284 amino acids and contains just the first DM14 domain ( Fig 2A ) . Homologous recombination of ES cells was analysed by Southern Blotting and subsequently functionality of loxP sites in ES cell clones was confirmed by a cell permeable His-TAT-NLS-Cre [38] prior to blastocyst injection . Cc2d1aneoflox mice were bred with Flp deleter and Cre deleter lines and Cc2d1a deficiency was confirmed by genotyping PCR , RT-PCR and Immunoblotting ( Fig 2B–2D ) . The truncated protein could not be detected . Heterozygous animals are viable , fertile and display no obvious abnormalities . Genotyping of litter from heterozygous breedings revealed no homozygous littermates , suggesting that no homozygous animal survive until weaning age ( 3–4 weeks ) . Closer examination of newborn pubs revealed that homozygous mice die within a few hours after delivery , similar to recently published studies [15 , 17 , 21] . Size and weight of the homozygous pups did not differ from their littermates but they displayed severe difficulties in breathing and turned cyanotic within a few minutes , suggesting that these pups suffer from respiratory distress ( see Fig 3C ) . An open question is whether a defect in the brain is responsible for the breathing defect that causes the perinatal death of Cc2d1a deficient mice [21] . This assumption could not be proven , as no conditional knockout was available . Our conditional knockout mouse model enabled us to test this hypothesis . To do so , we crossed Cc2d1aflox/flox mice with Nestin-Cretg/+ mice , which lead to brain specific inactivation of Cc2d1a [39] . Cre-mediated recombination of the Cc2d1aflox allele in CNS neurons and glia cells was confirmed by PCR analysis of DNA extracted from brain and tail biopsies of mutant and wild type embryos ( E14 . 5 ) and immunoblotting with CC2D1A antibodies ( Fig 3A and 3B ) . Extensive breeding of Nestin-Cretg/+ , Cc2d1aflox/+ animals with homozygous Cc2d1aflox/flox mice did not lead to any conditional Cc2d1a mutant animals at weaning age ( S1 Table ) suggesting early lethality of the Nestin-Cretg/+ , Cc2d1aflox/flox animals . As the genotype distribution of conditional Cc2d1a mutant embryos up to E18 . 5 was normal ( S2 Table ) we had a closer look at newborn animals and compared Cc2d1a deficient animals with Nestin-Cre conditional Cc2d1a mutants . We monitored five births for both lines . Mutant animals were born in expected Mendelian ratios ( S3 and S4 Tables ) . Eight out of eight Cc2d1a deficient animals failed to breathe and turned cyanotic after birth , while this was the case in only four out of twelve brain specific Cc2d1a mutants ( Fig 3C and 3D ) . Nevertheless , all brain specific Cc2d1a mutants died within 12 hours after birth , as no animal was observed the morning after giving birth probably due to the mother cannibalizing dead litter . The slight difference in the time of death between the full knockout animals and the brain specific mutants might suggest that the reason for the perinatal death is not entirely nervous system dependent or that the background of mice influences the time of death . The latter is supported by the fact that Cc2d1a deficient mice of a C57BL/6J background [17] die within one day after birth while Cc2d1a deficient mice of 129/Sv background die within minutes after birth [21] . As Cc2d1a mice in this study were derived from a C57BL/6N x 129Sv hybrid line backcrossed several generations to C57BL/6N and the Nestin-Cre animals are of a C57BL/6J genetic background , the observed difference might be strain-dependent . To test this possibility we crossed Cc2d1a+/- animals with C57BL/6J wild type or C57BL/6J-Nestin-Cre+/- mice to generate Cc2d1a mutants with the same background as the Nestin-Cre conditional mutants . We monitored three births with five Cc2d1a deficient animals ( S5 Table ) . Five out of five mutant animals displayed no breathing after birth and turned cyanotic like the mutants on the original background , suggesting that the analysed backgrounds do not influence the time of death . Taken together , these results indicate that the loss of CC2D1A in the nervous system is sufficient to induce the full knockout phenotype ( namely the perinatal death due to respiratory distress ) , even though with incomplete penetrance . CC2D1A was recently described to localise to centrosomes and regulate centriole cohesion in HeLa cells . siRNA mediated knockdown of CC2D1A in these cells was shown to cause formation of multipolar spindles accompanied by separase dependent centriole splitting [18] . To review these finding in murine cells we established murine embryonic fibroblast ( MEF ) cell lines isolated from Cc2d1a-/- , Cc2d1a+/- and wild type embryos and stained them with CC2D1A specific antibodies . The commercially available antibodies used in previous studies gave strong signals in our assay in all genotypes suggesting unspecific binding of the antibodies to other cellular antigens ( see S2 Fig ) . To generate a specific polyclonal CC2D1A antibody we immunized guinea pigs with a murine antigen that comprises the 4th DM14 domain ( aa 481 to 640 ) fused to the extended C-Terminus of CC2D1A ( aa 788 to 943 ) that is absent in CC2D1B . Immunocytochemical staining of MEF cells revealed a specific signal of CC2D1A in wild type MEFs , a weaker signal in heterozygous Cc2d1a+/- cells and no signal in homozygous Cc2d1a-/- cells ( Fig 4A ) , confirming the specificity of the antibody . It is reported that CC2D1A co-localises with the centrosome marker γ-Tubulin in HeLa cells and its depletion causes an increase in the mitotic index and a reduction in the number of viable cells [18] . To confirm this finding in primary murine cells we stained MEF cells with anti-γ-Tubulin and anti-CC2D1A ( Fig 4B ) . We did not find significant co-localisation of γ-Tubulin and CC2D1A . However , when we over-expressed human CC2D1A-dsRed we found co-localisation with γ-Tubulin in several cells ( Fig 4B ) , suggesting that CC2D1A proteins do localise close to the centrosomes upon overexpression . In contrast , human EGFP-CC2D1B does not co-localise with γ-Tubulin upon over-expression ( Fig 4B ) . We did not observe a significant difference in the mitotic indices of wild type ( 0 . 02% , 974 cells ) and Cc2d1a deficient ( 0 . 01% , 1478 cells ) MEFs ( Fig 4C ) . Finally , a cell proliferation assay did not reveal significant variation between genotypes ( Fig 4D ) . In summary , while CC2D1A appears to contribute to centrosome function in HeLa cells , we failed to find any evidence for this function in primary murine cells . In D . melanogaster Lgd was shown to be involved in the regulation of the endocytic trafficking of the Notch receptor and loss of its function results in uncontrolled ectopic activation of the pathway [1 , 3 , 4] . To test whether this function in regulation of Notch activity is conserved in mammals , we analysed mice that lack CC2D1A in the intestinal epithelium . The role of Notch signalling during maintenance of this tissue is well understood . To generate the epithelium specific Cc2d1a knockout , we bred Cc2d1aflox/flox mice with Villin-Cretg/+ transgenic mice [40] . Loss of CC2D1A was confirmed by Western blotting ( Fig 5A ) . Sections of the small intestine of adult Villin-Cretg/+ , Cc2d1aflox/flox , Cc2d1b-/- and control mice were stained with Nuclear fast red and Alcian blue to label Goblet cells , whose differentiation is dependent on Notch signalling [41] . We found that the overall morphology of the epithelia was similar in wild type and Cc2d1a and Cc2d1b mutant intestines . Moreover , no significant difference was observed in the number and distribution of goblet cells in the analysed genotypes ( Figs 5B and S3 ) . In addition , dividing cells , visualised by staining with the proliferating cell antigen Ki-67 , were restricted to the crypts of the small intestine in the Cc2d1a mutant and in the wild type ( Fig 5C ) and no ectopic proliferation of progenitor cells , as seen in animals that ectopically express the constitutively active form of Notch1 ( N1ic ) [35] could be detected . This indicates that also the Notch dependent proliferation of intestinal progenitor cells is not impaired upon loss of Cc2d1a function . To directly monitor Notch activity in the gut epithelium , we used the recently published Notch reporter line Hes1-EmGFPSAT [42] . In this knock-in line emerald-GFP is expressed under the control of the endogenous promoter of the Notch transcriptional target gene Hes1 . Comparing GFP signals in cryosections of wild type and Hes1-emGFP intestines revealed high background of fluorescence . To identify putative changes in the GFP expression we immunostained mutant and control intestines with anti-GFP antibodies . GFP was detected mainly in the crypts of the intestine , where Notch signalling is known to regulate stem cell self-renewal and progenitor cells . No obvious differences in the amount of GFP-expressing cells between the analysed genotypes were found ( Fig 5D ) . Furthermore , a quantitative analysis of Hes1 mRNA expression levels in mutant and control mice revealed no differences between animals ( Fig 5E ) suggesting that Notch signalling is not altered in mice lacking CC2D1A or CC2D1B in the intestine . In flies loss of Lgd function causes a general defect in endosomal trafficking that affects several cargo proteins . To analyse whether Notch degradation is impaired in Cc2d1a deficient MEF cells , we performed a pulse-chase antibody uptake assay on wild type and Cc2d1a deficient MEFs ( Fig 6A ) . For this purpose MEF cells were stably transduced with human NOTCH1 tagged with an HA epitope in its extracellular domain [43] . The generated cell lines display comparable NOTCH1-HA expression levels ( Fig 6C ) . For the assay , they were incubated with an Alexa488-labelled anti-HA antibody at 4°C and then shifted for various periods of time to 37°C ( see also [44] ) . Initially , NOTCH1 was detected at the cell surface and after a 30 min chase NOTCH1 was also visible in intracellular vesicles in both analysed genotypes . After 60 min the receptor was rarely located at the cell surface but in intracellular vesicles . The number of labelled cells decreased constantly , with only very few cells labelled after 180 min of incubation ( at 37°C ) . Calculating the number of fluorescent labelled cells in 2 independent experiments revealed no striking differences between wild type and Cc2d1a deficient cells ( Fig 6B ) , suggesting that NOTCH1 endocytosis and degradation is not affected by the lack of CC2D1A . To determine the subcellular localisation of CC2D1A , we stained wild type MEFs using our specific CC2D1A antibody together with organelle markers ( S4 Fig ) . We found that CC2D1A is distributed in punctate structures in the cytosol . We could not associate these punctae with distinct organelles , indicating that CC2D1A is not localised on the ER , the Golgi or endosomes but accumulate in unidentified punctae . Moreover , we found little co-localisation of CC2D1A and CHMP4B and neither of them could be detected on RAB7 positive late endosomes ( Fig 7A and 7A` ) . CC2D1 proteins have a C2 binding domain that has been shown to be able to bind phospholipids present in endosomal membranes [4] . To test whether CC2D1A and CC2D1B can associate with membranes in general , we performed a cell fractionation assay . We detected CC2D1A and CC2D1B in the cytosolic as well as in the membrane fraction , in contrast to control proteins , which were correctly restricted to only one fraction ( S5A Fig ) . This result indicates that a fraction of CC2D1A and also CC2D1B is associated with membranes . CHMP4 proteins cycle between the endosomal membrane and the cytosol . An endosomal localisation of the CHMP4 yeast ortholog Snf7 could only be observed in the absence of Vps4p function [28] . Loss of Vps4p leads to an accumulation of ESCRT-III on endosomal membranes and disrupts proper ILV formation . We wondered whether this is also the case for CHMP4 proteins in mammals and , since CC2D1 proteins physically interact with CHMP4 proteins ( see below ) , whether CC2D1A also accumulates at the endosomal membrane upon reduction of VPS4 function . The mammalian genome contains two Vps4 genes , Vps4a and Vps4b [45] . Homozygous Vps4a-/- and Vps4b-/- mice die during early embryonic stages . To analyse the influence of the two mammalian orthologs VPS4A and VPS4B on subcellular localisation of CC2D1A , we generated MEFs that are double heterozygous for Vps4a and Vps4b . Strikingly , we observed dramatically enlarged LAMP1 positive MEs/lysosomes in these cells , although the corresponding mice were healthy and displayed no detectable phenotype ( Fig 7B ) . Moreover , CHMP4B accumulated on the enlarged MEs of the double heterozygous cells ( Fig 7B and 7B` ) . This indicates that a reduction of VPS4 function results in the accumulation of CHMP4 proteins on the endosomal membrane also in mammals . Importantly , also CC2D1A accumulated on these CHMP4B/LAMP1 positive vesicles ( Fig 7B and 7B` ) . In a complementary experiment , we over-expressed a dominant negative VPS4B ( VPS4BE235Q ) that lacks its ATPase activity leading to aberrant endosomal structures [46] . Indeed , we could confirm that expression of VPS4BE235Q in wild type MEFs leads to the formation of enlarged endosomal structures while expression of normal VPS4B did not lead to any changes ( S5B and S5B` Fig ) . Moreover , FK2 , a marker for ubiquitinated proteins , strongly labelled the VPS4BE235Q induced aberrant MEs , suggesting that the removal of CHMP4 from the endosomal membrane is impaired . As in Vps4a+/- , Vps4b+/- cells , CHMP4A and CC2D1A accumulated on the enlarged MEs ( S5B and S5B` Fig ) . These results indicate that CC2D1A might function at the endosome and cycles between the cytosol and the endosomal membrane , just like its interaction partner CHMP4B . They also indicate that CC2D1A appears to be removed from the endosomal membrane by VPS4 and suggest that CC2D1A is involved in ESCRT-III function in mammals . Unfortunately , we were unable to confirm an endogenous endosomal localisation for CC2D1B , as our antibody was not specific in histological staining . Nevertheless , over-expressed CC2D1B also accumulated on the enlarged vesicles in Vps4a+/- , Vps4b+/- cells ( S5C Fig ) , suggesting that CC2D1A and CC2D1B are temporally associated with the endosomal membranes and are involved in ESCRT function . So far only in vitro data showed physical interactions of CHMP4B with CC2D1A . In order to test whether this can also be observed between the corresponding endogenous proteins in a cell , we performed the proximity ligation assay ( PLA ) . PLA enables the in situ detection of endogenous protein protein interactions , using protein specific antibodies [47] . For the assay , we compared wild type and Cc2d1a deficient MEF cells and used our specific CC2D1A antibody in combination with a commercially available CHMP4B antibody . As expected , only few PLA signals were visible in Cc2d1a deficient cells ( 0 . 6±1 . 07 ) , confirming the accuracy of the method ( Fig 8 ) . In contrast , abundant PLA signals were detected in wild type MEFs ( 6 . 47±3 . 18 ) indicating that endogenous CC2D1A and CHMP4B interact ( Fig 8 ) . To exclude the possibility that CHMP4B and CC2D1A interact in the cytosol randomly by freely diffusing cytosolic proteins we performed PLA with CHMP4B and the cytosolic protein PGK1 ( S6 Fig ) . Again , we could detect a clear interaction only in the control and not between CHMP4B and PGK1 ( 5 . 61±2 . 39 for CHMP4B with CC2D1A compared to 0 . 64± 1 . 01 for CHMP4B with PGK1 ) . Taken together , these experiments further support the assumption that CC2D1A and possibly also CC2D1B are involved in ESCRT-III mediated events . To further elucidate a function of CC2D1 proteins in endosome maturation we analysed the endo/lysosomal compartments of wild type and mutant MEFs with the transmission electron microscope ( TEM ) ( Fig 9 ) . Endo/lysosomal organelles were identified by their ultrastructural characteristics proposed by others [48–50] . Exemplary pictures of the analysed endo/lysosomal organelles are depicted in S7A–S7D Fig . Cc2d1b-/- cells show no differences in size or morphology of the endo/lysosomal compartments compared to wild type cells ( Figs 9A–9A” , 9C–9C” , 9E and S7E ) . In contrast and in accordance with the fluorescence microscopy data , Vps4a+/- , Vps4b+/- cells had enlarged endo/lysosomal compartments . We detected large multi-membrane structures that resembled the class E compartment described for loss of ESCRT function in yeast and mammalian cells [51 , 52] . We did not observe these structures in Cc2d1a-/- cells , but found that the average size of their endo/lysosomal organelles was significantly increased ( Figs 9B–9B” , 9E and S7E ) . The endo/lysosomal phenotype of Cc2d1a-/- cells resembled that observed in lgd mutant cells of D . melanogaster [53] . When we allocated the endo/lysosomal organelles to different size classes , we found classes in Cc2d1a-/- and Vps4a+/- , Vps4b+/- MEFs that were not present in Cc2d1b-/- or wild type MEFs ( S7E Fig ) . Thus , the loss of Cc2d1a function results in a moderate enlargement of the endo/lysosomal compartment , indicating that it is involved in endosomal maturation . Moreover , already the double heterozygosity of Vps4a and Vps4b results in a dramatic morphological defect of the endo/lysosomal phenotype , which appears to have no direct consequences for the viability of mice in captivity . In order to obtain further evidence for the conservation of the function of the Lgd proteins in metazoans , we asked whether and to what extent the two mammalian orthologs of Lgd could replace the loss of function of lgd in D . melanogaster . Since it turned out that rescue experiments with the Gal4/UAS system are not suitable [22] , our previously reported results are not meaningful [1] . We therefore here used an assay that we have later developed for the structure-function analysis of Lgd [22] . In short , human CC2D1A and CC2D1B were cloned behind the lgd promoter ( lgdP ) and the constructs were inserted into the same landing site to neutralise position effects on expression . Thus , CC2D1A and CC2D1B are expressed at endogenous level and similar to each other . This allowed the direct comparison of the effects caused by the two constructs . Loss of lgd function in D . melanogaster results in over-proliferation of the imaginal discs , which are epithelial monolayers ( Fig 10A and 10B ) . Moreover , target genes of Notch such as wingless ( wg ) are ectopically expressed , indicating that the Notch pathway is ectopically activated ( [2] , Fig 10A and 10B ) . We found that both orthologs could rescue the lgd mutant phenotype , albeit to different extent ( Fig 10C–10F ) . lgdP-CC2D1B completely rescued the mutant , even if only one copy was present in the genome ( Fig 10C ) . We obtained normal looking fertile flies . In the case of lgdP-CC2D1A only two copies present resulted in a complete rescue of the imaginal disc phenotype ( Fig 10D–10F ) . However , in this case the fully differentiated flies failed to hatch . These experiments reveal functional differences between the human CC2D1 proteins for the first time and indicate that CC2D1B is more similar to Lgd than CC2D1A . Nevertheless , both orthologs can rescue the lgd mutant phenotype to different extent , indicating a significant overlap in their function . The rescue of lgd mutants with only one copy of lgdP-CC2D1A resembled that of weak lgd mutants [1] . We have previously shown that the phenotype of lgd mutants dramatically worsened if one copy of shrb is additionally removed ( shrb heterozygosity ) [22] , indicating a functional relationship between the two loci ( Fig 10H ) . We used this test of genetic interaction between shrb and lgd to test whether also CC2D1A requires the interaction with Shrb in D . melanogaster . We found that the phenotype of lgdd7;lgdP-CC2D1A also dramatically worsened if one copy of shrb was removed ( Fig 10G; genotype: lgdd7 shrb4-1/lgdd7; lgdP-CC2D1A/+ ) . This genetic interaction suggests that Shrb functionally interacts with CC2D1A in a similar manner as we have found for Lgd [22] . Our results suggest for the first time that the CC2D1 proteins functionally interact with members of the CHMP4 family in an organism .
Here we report the first characterisation of the Lgd homologue CC2D1B in a mammalian system . We found that it is not an essential gene in mouse , since its loss of function results in no detectable phenotype . We confirmed the previously reported finding that the loss of function of its paralogue CC2D1A results in postnatal death due to failure to breathe [15 , 17 , 21] . We here extend this finding by showing that the postnatal death is mostly caused by loss of its function in the nervous system . This result extends previous studies , that suggested either an involvement of CC2D1A in neuronal differentiation and brain development via the regulation of the Protein Kinase A [15] or in functional maturation of synapses [21] . Thus , a defect in certain neurons or glia cells is probably the cause of the breathing defect and the resulting death of mutant animals . We failed to find evidence for an involvement of the CC2D1 proteins in cell division . Instead Cc2d1a or Cc2d1b mutants develop normally and MEFs harvested from these homozygous mutants do not show a cell division phenotype nor are they multi-nucleated as has been found for HeLa cells treated with CC2D1A specific siRNA [18] . This is in agreement with the fact that mutant animals develop until birth with no obvious defects . In D . melanogaster the loss of function of lgd results in the constitutive ligand-independent activation of Notch in several epithelia [1–4 , 6] . In order to test whether the loss of one of the orthologs results in ectopic Notch activation in mammals , we analysed their function in the epithelium of the gut of mice . In contrast to D . melanogaster , we did not find any evidence for ectopic activation of the Notch signalling pathway if one of the two orthologs is inactivated . The number or distribution of goblet cells did not change and the expression of the Notch activity reporter Hes1-emGFP and endogenous Hes1 was unaffected upon loss of function of Cc2d1a or Cc2d1b . This suggests that the individual loss of the genes does not result in uncontrolled activation of the Notch pathway in mammals . Notch signalling is critical for T cell development ( reviewed by [54] ) . In Cc2d1a deficient embryos the development and function of thymocytes is not impaired [21] further supporting the notion that Notch signalling is not altered in the absence of Cc2d1a alone . It might still be possible that weak ectopic activation of the Notch pathway occurs in mutant mice , but this activation is too weak to be detected with our test systems and to manifest itself in phenotypes in the gut epithelium . However , it is possible that this weak activity has effects in other tissues , such as the brain . Indeed , it has been reported that CC2D1A over-expression leads to weak Notch reporter activity in HEK293 cells [20] . This weak activation might be responsible for the brain defect , since the Notch pathway is known to regulate neural development and postnatal neurogenesis ( reviewed by [55] ) . Moreover , activation of the pathway in cultured neurons changes their dendritic arborisation [56 , 57] . Importantly , a similar arborisation defect together with activation of the Notch pathway has been observed upon inactivation of Shrub , the ortholog of the CHMP4s in D . melanogaster [27] . However , CC2D1A was also shown to act as a positive regulator of the cAMP/PKA pathway [14 , 15] that is known to regulate synaptic plasticity , learning and memory [58 , 59] and the NF-κB pathway [7] that is involved in the regulation of neuronal differentiation and survival [60] . Intriguingly , NF-κB activation was shown to be specific to CC2D1A , while CC2D1B could not activate NF-κB [20] suggesting that the proteins are not functional redundant with regard to NF-κB signalling . These finding could also explain why Cc2d1a deficient mice die early while Cc2d1b deficient mice display no obvious phenotype . Our pulse and chase experiment to analyse the endosomal degradation of the human NOTCH1 receptor showed no alterations in wild type and Cc2d1a deficient MEFs . Nevertheless , we cannot rule out the possibility that our method might not be sensitive enough for detection if a slight delay or defect in degradation occurs . The siRNA mediated depletion of CC2D1A in HeLa cells , analysed with high-resolution confocal microscopy and quantitative multiparametric image analysis , detected a decrease in EGF and TF endocytosis [29] , supporting a function of CC2D1A in the endocytic pathway . Indeed we found clear evidence for an involvement of CC2D1 proteins in the endocytic pathway: 1 . Our EM analysis revealed that the endosomes are enlarged upon loss of Cc2d1a function . Importantly , the classification according to size revealed that larger size classes appear in the mutant that are absent in the wild type . This has also been found lgd mutant cells in D . melanogaster [53] . 2 . We found that a fraction of both CC2D1 proteins is associated with membranes . 3 . While CC2D1A is distributed in the cytosol in wild type MEFs , it accumulates on endosomal membranes of Vps4a+/- , Vps4b+/- double heterozygous MEFs . This relocation is striking and suggests that , like the CHMP4 proteins , CC2D1A and probably also CC2D1B cycles between the cytosol and the limiting membrane of the endosome . It also suggests that both proteins are involved in the function of ESCRT-III . 4 . We show that CC2D1A interacts with CHMP4B at endogenous level . This has not been shown for the full-length proteins and especially not under endogenous conditions . Thus , our findings show for the first time that this interaction occurs in vivo . 5 . We show that both CC2D1 proteins can replace the function of Lgd in D . melanogaster . In the fly Lgd is evidently involved in endosomal trafficking of transmembrane proteins and both Lgd orthologs can take over this function . Moreover , CC2D1A genetically interacts with the CHMP4 ortholog Shrub in D . melanogaster , suggesting that it requires Shrub respectively the CHMP4 orthologs for its function . Interestingly , the rescue ability of both orthologs differed: While the imaginal disc phenotype could be completely rescued by both CC2D1 proteins , only flies expressing CC2D1B were viable and fertile . CC2D1A contains an extended C-Terminus with no putative domains that is missing in Lgd and CC2D1B . This C-terminal extension might be the reason for the poorer rescue ability and might be required for the suggested additional non-redundant functions of CC2D1A in mammals . Despite the differences in the extent of the rescue , our experiments in D . melanogaster suggest that CC2D1A and CC2D1B have overlapping functions in the endocytic pathway . Thus , it is likely that functional redundancy among both CC2D1 proteins prevents the formation of a detectable cellular phenotype if only one gene is deleted , especially since we found that both proteins are present in all cells tested . Therefore , it will be necessary to delete both genes to further evaluate their function in the endocytic pathway and Notch pathway regulation . The analysis of these double mutants will illuminate the function of CC2D1A and CC2D1B in mammals .
The Cc2d1a targeting vector was generated using the pGK12 vector ( Artemis Pharmaceuticals , Köln ) which contains a PGK neomycin cassette , as a positive selection marker , and the Herpes Simplex virus thymidine kinase ( TK ) gene , as a negative selection marker . The 2 . 3 kb 5`arm , the deleted area ( 4kb , exons 7–14 ) and the 5 . 0 kb 3`arm were amplified via PCR using BAC-DNA RP23-298K21 ( BACPAC Resources , Oakland , USA ) as a template and confirmed by sequencing . The 129/B6 F1 hybrid ES cell line v6 . 5 [61] was electroporated with the NotI linearized targeting vector . Using Southern Blot analysis five targeted ES cell clones were identified and injected into blastocysts of CB20 mice and transplanted into pseudopregnant host mothers . The resulting chimeric mice were bred to C57BL/6N mice . Resulting Cc2d1aneoflox/+ animals were crossed with FLPe-Deleter mice [62] to delete the FRT-flanked neomycin resistance cassette ( Cc2d1aflox/+ mice ) . Heterozygous Cc2d1aflox/+ siblings were crossed to generate homozygous Cc2d1aflox/flox mice that are viable , fertile and do not display any obvious phenotype . For a full knockout Cc2d1aneoflox/+ animals were crossed with Cre-Deleter mice [63] . Resulting Cc2d1a+/- siblings were crossed to generate homozygous Cc2d1a-/- mice . The mouse strain C57BL/6N-Cc2d1btm1a ( KOMP ) Wtsi used for this research project was created from ES cell clone EPD0017_2_C11 obtained from the NCRR-NIH supported KOMP Repository ( https://www . komp . org/ ) and generated by the CSD consortium for the NIH funded Knockout Mouse Project ( KOMP ) . Methods used on the CSD targeted alleles have been published elsewhere [64] . Cc2d1btm1a ( KOMP ) Wtsi mice were crossed with C57BL/6-TgN ( FLPe ) mice [62] to obtain the conditional allele Cc2d1bflox . Cc2d1bflox/flox animals do not display any obvious phenotype and are fertile . For a complete knockout , Cc2d1bflox/flox animals were crossed with C57BL/6;129/Ola-TgH ( Deleter+8 ) KP animals [63] to delete the floxed exon to generate a frameshift mutation . Resulting Cc2d1b+/- siblings were crossed to generate homozygous Cc2d1b-/- mice . The C57BL/Vps4a+/- heterozygous knockout mouse strain resulted from a conditional Vps4a knockout model . Vector construction and targeted knockout strategy was designed together with genOway ( Lyon , France ) , where mice were generated . 129Sv/Pas ES cells were transfected with the targeting construct pVPS4a-KOE1-HR . This vector contains a short ( 1 , 9 kb ) and a long ( 5 , 8 kb ) Vps4a homology region . Two LoxP sites were inserted flanking Vps4a exon 2 and exon 3 . The positive selection neomycin gene flanked by FRT sites is inserted upstream of exon 2 and the negative selection marker Diphteria Toxin A ( DTA ) is located downstream from the distal long homology arm at the 3’ end of Vps4a . Homologous recombination in ES cells was demonstrated by PCR and Southern blot analysis . After blastocyst injection male chimeras were mated with female Flp-expressing mice ( C57BL/6J ) to excise the neomycin selection cassette and to generate heterozygous mice carrying the floxed Vps4a allele . Breeding floxed Vps4a mice with C57BL/6J Cre-expressing mice led to constitutive heterozygous Vps4a knockout mice lacking exon 2 and 3 . The heterozygous Vps4a knockout mice were inbred with C57BL/6 mice for more than 10 generations . The 129Sv/Vps4b+/- is a classical knockout mouse strain that was generated using the targeting construct pVPS4b-KO , where the Vps4b exon 1–4 region is replaced by a neomycin-resistance cassette . The negative selection marker DTA is located at the 3’ end of the Vps4b homology region . Following homologous recombination in ES cells and blastocyst injection , heterozygous Vps4b knockout mice were confirmed by Southern blot analysis and inbred with 129Sv mice for more than 10 generations . This Vps4b knockout mouse strain was generated in collaboration with U . Rüther ( Düsseldorf ) . A more detailed description of the experimental steps will be published elsewhere together with a comprehensive phenotype analysis of conditional and constitutive Vps4a and Vps4b mice . Nestin-Cre mice ( C57BL/6J . B6SJF2-TgN ( Nestin-Cre ) ) have been described elsewhere [39] and were a kind gift from R . Kühn . Villin-Cre mice ( B6 . SJL-Tg ( Vil-cre ) 997Gum/J ) were purchased from The Jackson Laboratory ( stock number 004586 ) . Hes1-emGFPSAT mice were a kind gift from S . Fre [42] . C57BL/6-TgN ( FLPe ) mice have been described elsewhere [62] and were a kind gift of A . Gödecke . C57BL/6;129/Ola-TgH ( Deleter+8 ) KP mice have been described elsewhere [63] and were a kind gift of K . Pfeffer . Mice were maintained in the central animal research facility of the Heinrich-Heine-University Duesseldorf under specific pathogen-free conditions . Animals got food and water ad libitum . All experiments performed on animals in this study were approved by the Animal Care and Use Committee of the local Government of Düsseldorf in accordance with the German law for animal protection and were carried out with the authorization of the LANUV ( Landesamt für Natur- , Umwelt- und Verbraucherschutz ) of North-Rhine-Westphalia , Germany under the reference number 9 . 93 . 2 . 10 . 31 . 07 . 249 . Genomic DNA was isolated from tail biopsis using DirectPCR Tail buffer ( Viagen ) and PCR was carried out using 1 unit of Crimson Taq DNA polymerase ( NEB ) in a 25 μl standard reaction mix . Oligonucleotide sequences for genotyping are as follows: Cc2d1a: P1: 5`agaccctgtggctggattgt3` , P2: 5`acccatcctttgcttgtctc3` , P3: 5`gccagcctggtctacaatca3`and P5: 5`cctgacctgagtactggaca3` . Cc2d1b: 1F: 5`gcatgtgccacaatgccaagc3` , 3R: 5`ctgagtgagcagttcctagc3`and 3bR: 5`aggctgcctctaagggttcc3` . Nestin-Cre: p-CRE1: 5`atgcccaagaagaagaggaaggt3` , pCRE2: 5`gaaatcagtgcgttcgaacgctaga3` , oIMR7338: 5`ctaggccacagaattgaaagatct3`and oiMR7339: 5`gtaggtggaaattctagcatcatcc3` . Villin-Cre: P1878: 5`gtgtgggacagagaacaaacc3` , P1879: 5`acatcttcaggttctgcggg3` , P8744: 5`caaatgttgcttgtctggtg3`and P8745: 5`gtcagtcgagtgcacagttt3` . Hes1-emGFPSAT: P35: 5`cccaagttcgggtgaaggc3`and P36: 5`ccttggacaatgccacccaa3` . Vps4a: Vps4a-5F: 5`tataatatggttgagcctcccttc3` , 3 . 1 Rev Vps4a: 5`gcaccccaaactggaaaaccacttactctcc3` , Vps4a-5R: 5`attcgtgacctatctcgattcttc3` . Vps4b: Vps4bKO_NEO_For1: 5`aggattgggaagacaatagcag3` , Vps4b_WT_For2: 5`tgctttgaggaactaaatcatcc3` , Vps4b_WT_Rev2: 5`ggattggactcaatgcctacat3` . For the production of a CC2D1B/mLGD1 antibody a murine Cc2d1b cDNA fragment ( covering amino acids 1–253 ) was amplified from RIKEN Mouse FANTOM Klon A830039804 with the primers mLGD1-AK-For ( 5`aaactcgaggcatgccagggccaagacc3` ) and mLGD1-AK-Rev ( 5`aaagcggccgctggctccatggcacaggga3` ) and cloned into the GST affinity tag containing vector pGEX-6P-2 ( GE Healthcare Bio-Sciences ) using XhoI and NotI restriction sites . The GST-CC2D1B ( 1–253 ) fusion protein was expressed in E . coli and captured by immobilized glutathione . After removal of GST affinity tag by PreScission Protease guinea pigs were immunized with the CC2D1B ( 1–253 ) antigen ( Cocalico Biologicals Inc . ) . Serum from the final bleed was further affinity purified with the antigen immobilized on nitrocellulose membrane . For the production of a CC2D1A/mLGD2 antibody guinea pigs were immunized ( Eurogentec s . a . ) with a GST tagged antigen that consists of the 4th DM14 domain ( amino acids 481 to 640 ) fused to the extended C-Terminus of CC2D1A ( amino acids 788 to 943 ) that is absent in CC2D1B . For this two cDNA fragments ( 477–647 and 788–943 ) were amplified using the primer pairs mLGD2-481-For ( 5`aggatccaacagagcccagcagcagct3` ) /mLGD2-640-Rev ( 5`acccgggcttgatgacactgaaggtcctctg3` ) and mLGD2-788-For ( 5`acccggggcccagcagttggaaactacaac3` ) /mLGD2-943-Rev ( 5`actcgagtcacctgcggagtcgctgca3` ) and cloned into pGEM-T Easy vector ( Promega ) . Ligation of SmaI and DraIII digested fragments led to fusion of both cDNA fragments . Using BamHI and XhoI digestion the cDNA was cloned into the GST affinity tag containing vector pGEX-6P-2 ( GE Healthcare ) . The GST-CC2D1A ( 477–647 , 788–943 ) fusion protein was expressed and purified by immobilized glutathione before the immunization of guinea pigs ( Eurogentec s . a . ) . Serum from the final bleed was further affinity purified with the antigen immobilized on nitrocellulose membrane . After homogenisation and short sonication of tissue in lysis buffer ( 20 mM Tris-HCl pH 7 . 4 , 100 mM NaCl , 10% Glycerol , 0 . 5% Triton X-100 , proteinase inhibitor cocktail ( Roche , 1:200 ) and 1 mM PMSF ) , the lysates were centrifuged at 20 , 000 x g for at least 20 minutes at 4°C . The soluble fraction was collected and loaded onto 8–10% SDS-PAGE gel and immunoblotted according to standard protocols . The following antibodies were used: Actin ( Sigma , A-5060 , 1:5000 ) , alpha-Tubulin ( Sigma-Aldrich , Clone B-5-1-2 , 1:10000 ) , CC2D1A ( Abnova , H00054862-BO1P , 1:2000 ) , CC2D1B ( Proteintech , 1:4000 ) , CHMP4B ( Santa Cruz C12 , 1:2000 ) , LAMP1 ( DSHB , 1D4B , 1:2000 ) and PGK1 ( Gene Tex , GTX107614 , 1:500 ) . For the quantification of protein levels CC2D1A/mLGD2 ( this study , 1:500 ) and CC2D1B/mLGD1 ( this study , 1:1000 ) antibodies were used . Relative protein levels were determined using ImageJ . Total RNA was isolated from tissues using TriFast Reagent ( Peqlab ) . 1–3 μg of isolated RNA was treated with DNase I and then reverse-transcribed using the SuperScript First-Strand Synthesis System ( Invitrogen by Life Technologies ) or the GoScript Reverse Transcription System ( Promega ) according to manufacturers`instructions . RT-PCR was carried out with 0 . 5 μl of cDNA , 0 . 2 μM each of forward and reverse primers , 200 μM dNTPs and 0 . 5 units of Crimson Taq DNA polymerase ( NEB ) in a 25 μl reaction . Cycling conditions were as follows: 35 cycles of 94°C ( 30 s ) , 56°C ( 30 s ) , and 72°C ( 30 s ) . mHPRT ( hypoxanthin-phosphoribosyl-transferase ) was used as internal control . Oligonucleotide sequences for RT-PCR are as follows: mHPRT-for: 5`cgtcgtgattagcgatgatg3` , mHPRT-rev: 5`tatgtcccccgttgactgat3` . CC2D1A-Ex2-F: 5`gggattaatgaggaggagctg3` , CC2D1A-Ex4-R: 5`gccttctgttcttctccaagg3` , CC2D1A-Ex12-F: 5`tagtgggtgtcctggaaactg3` , CC2D1A-Ex13-R: 5`gtcctgatggaggtgctttg3` , CC2D1A-Ex22-F: 5`agatcctggaggttttggatg3` , CC2D1A-Ex25-R: 5`caacacgctaaggctatgcag3` , CC2D1B-Ex1-F: 5'gagaaggcccggttttggtt3' , CC2D1B-Ex4-R: 5'tacagtctgctgccagcttc3' , CC2D1B-Ex16-F: 5'aggaggtgtatgcccagcta3' , CC2D1B-Ex17-R: 5'cgagtggtctcagccacatt3' , CC2D1B-Ex24-F: 5'ctgagaactggctggtcctg3' and CC2D1B-Ex25-R: 5'acagagcagtggggtatcct3' . Quantitative analysis of Hes1 gene expression in the murine intestinal epithelia was performed by quantitative RT-PCR using a Stratagene Mx3005P cycler ( Agilent Technologies ) , with KAPA SYBR FAST qPCR Kit Master Mix ( 2x ) Universal ( Peqlab ) using 1 μl cDNA according to manufacturer´s instructions . Gapdh ( glyceraldehyde-3-phosphate dehydrogenase ) expression was used as an internal control . Oligonucleotide sequences for quantitative RT-PCR were as follows: mHes1-F: 5`accccagccagtgtcaaca3` , mHes-R: 5`tgtgctcagaggccgtctt3` , mGapdh-F: 5`tgaaggtcggtgtgaacgg3`and mGapdh-R: 5`cgtgagtggagtcatactggaa3` . The relative expression levels of Hes1 mRNA in the intestinal epithelia were calculated and analyzed by the 2-ΔΔCT method . Statistical analysis was performed using Student`s t-test , unpaired , two-tailed . For cryopreservation tissues were fixed in Stefanini`s Fixative ( 2% PFA , 0 . 2% Picric acid ) over night , treated with 30% Sucrose solution ( up to 12 hours ) and were embedded in Tissue Tek ( Sakura Finetek ) . Embedded tissues were cut on a cryostat into 5–7 μm thick slices and transferred to Superfrost slides . For antigen retrival sections were boiled in 10 mM citrate buffer ( pH 6 . 0 ) for 10 min . Sections were then washed 3 times in dPBS , permeabilised for 10 min in 0 . 3% Triton X-100 in dPBS and then incubated with a 10% blocking solution ( 10% NGS , 0 . 1% Tween-20 , 0 . 3 M Glycine in dPBS ) for 1 hour at room temperature . Sections were incubated with primary antibodies ( Ki67 ( abcam , ab16667 , 1:200 ) and GFP ( Roche , 11814460001 , 1:500 ) in a 1 . 5% blocking solution ( 1 . 5% NGS , 0 . 1% Tween-20 in dPBS ) overnight at 4°C . Sections were then rinsed three times in dPBS and subsequently incubated with secondary antibodies ( 1: 500 in 1 . 5% blocking solution ) for 1 hour at room temperature . After that , sections were washed in dPBS , the nuclei were counterstained with DAPI ( 0 . 3 μg/ml ) and mounted with VECTASHIELD ( Vector Laboratories ) . Alcian blue staining was performed according to standard protocols on paraffin sections . For the quantification of goblet cell number , the number of stained cells in 250μm of a crypt-villus unit ( 35–46 units/animal , n = 3 ) were counted . Statistical analysis was performed using Student`s t-test , unpaired , two-tailed . Human CC2D1A cDNA was amplified from clone 6585236 ( Source BioScience ) with the oligonucleotides dsRed-CC2D1A-For ( 5`aactcgagtggaattcgccatgcacaag3` ) and dsRed-CC2D1A-Rev ( 5`aaaagcttagcgtaatctggcacatcg3` ) and cloned into pDsRed-Monomer-N1 vector ( Clontech ) using HindIII and XhoI restriction sites . Human CC2D1B cDNA was synthesised ( from Ensembl transcript ENST00000284376 , GenScript ) and was amplified with the oligonucleotides EGFP-CC2D1B-For ( 5`aagaattcgcggcggcccatgatgc3` ) and EGFP-CC2D1B-Rev ( 5`atctagcatgctcgagtc3` ) and cloned into pEGFP-C1 ( Clontech ) using XhoI and EcoRI restriction sites . VPS4B cDNA was amplified from HeLa cDNA using the oligonucleotides VPS4-For ( 5`agaattcatgtcatccacttcgcccaacc3` ) and VPS4-Rev ( 5`agcggccgcgccttcttgaccaaaatcttc3` ) and cloned into pcDNA3 vector with an N-terminal Myc Tag ( Invitrogen ) using EcoRI and NotI restriction sites . The dominant negative VPS4B was generated by a site-directed mutagenesis PCR using the oligonucleotides VPS4B-E235Q-For ( 5`ccctccattatcttcattgatcaaattgattctc3` ) and VPS4B-E235Q-Rev ( 5`ccacagagagaatcaatttgatcaatgaagataatggaggg3` ) and pcDNA3-VPS4B-Myc as a template according to standard protocols . MEFs were isolated from E13 . 5-E15 . 5 mouse embryos according to standard protocols . After 2–3 passages cells were transfected with pMSSVLT SV40 vector containing the SV40 Large T Antigen and immortalized cells were selected by G418 treatment . HeLa and MEFs were maintained in DMEM supplemented with 10% FCS , Pen/Strep ( 50 Units/ml ) and , for immortalized cells with G418 ( 0 . 5 μg/ml ) , and cultured at 37°C in a humidified incubator with 5% CO2 . All transfections were done with Lipofectamine 2000 ( Life Technologies ) according to the manufacturer`s instructions . The cell proliferation assay was performed with CellTiter 96 AQueous Non-Radioactive Cell Proliferation Assay ( Promega ) according to the manufacturer`s instructions . For the calculation of the mitotic index 1000–1500 cells of two to three different passages per genotype were counted and the number of cells undergoing mitosis ( visualized by DAPI staining ) was divided by total cell number and multiplied by 100 . Cells ( three 10 cm dishes ) were cultured to confluence , washed in dPBS , harvested by scraping and centrifuged at 600 x g for 5 min and homogenised in 1 ml buffer containing 20 mM HEPES-KOH , pH 7 . 2 , 400 mM sucrose and 1 mM EDTA by passing through a 22 G needle . Homogenates were centrifuged at 1 , 000 x g for 3 min at 4°C to pellet cell debris and nuclei . Supernatants were centrifuged at 100 , 000 x g ( Ti 70 . 1 rotor , Beckmann ) for 45 min at 4°C . The supernatant contained the cytoplasm fraction and the pellet the membrane fraction , which was resuspended in 500 μl homogenisation buffer . The protein concentration of each fraction was determined via Bradford assay and equal amounts of protein were used for Western blot analysis . HA-tagged NOTCH1 expressing MEFs were generated by retroviral transduction according to standard protocols . Briefly , Plat-E ecotropic packaging cell line was transfected with pBABE vector containing human NOTCH1-HA ( kind gift of J . C . Aster [43] ) and virus supernatant was then used to infect MEFs . Infected cells were identified by puromycin selection and clonal populations were obtained by dilution . For the uptake assay cells were briefly washed in serum-free DMEM and then incubated at 4°C for 30 min with anti-HA-Alexa 488 ( Life Technologies ) diluted in DMEM ( 1:300 ) followed by brief washing in cold dPBS and incubation in serum-free DMEM for various periods of time at 37°C . Cells were then washed in cold dPBS and fixed with 4% PFA . For immunocytochemistry cells were plated on coverslips 24 hours before the experiment . The cells were fixed for 10 min in 4% cold PFA , washed in dPBS and permabilised in 0 . 3% Triton X-100 ( in dPBS ) for 10 min at room temperature . After washing and incubation in blocking buffer ( 0 . 1% Tween-20 , 10% NGS ( normal goat serum ) , 0 . 3 M Gycine in dPBS ) for at least 30 min , cells were incubated with primary antibodies in staining buffer ( 0 . 1% Tween-20 , 1 . 5% NGS in dPBS ) overnight at 4°C . The following day , the cells were washed in dPBS and incubated with secondary antibodies ( 1:500 diluted in staining buffer ) for an hour at room temperature . Subsequently , after another washing step the nuclei were stained with DAPI ( 3μg/mL ) and mounted with VECTASHIELD ( Vector Laboratories ) on microscope slides . Next , the cells were analysed using an Apotome microscope ( Axio Imager Z1m , Zeiss ) . Representative images were processed and brightness of images was uniformly adjusted to enhance contrast using Adobe Photoshop . For colocalisation studies the AxioVision Colocalization Module was used . The following antibodies were used in a 1:200–500 dilution: Calnexin ( Cell Signaling Technology , 2433 ) , CC2D1A/mLGD2 ( this study ) , CHMP4B ( Santa Cruz , C12 ) , LAMP1 ( DSHB , 1D4B ) , RAB5 ( abcam , ab18211 ) , RAB7 ( abcam , ab50533 ) , SYNTAXIN 6 ( Cell Signaling Technology , 2869 ) , γ-Tubulin ( Sigma , T5326 ) . CC2D1A antibodies that gave unspecific signals in Cc2d1a deficient MEFs were the following: Bethyl Lab ( CC2D1A , A300-285A ) , Santa Cruz Biotechnology ( Freud-1 ( K19 ) , sc-79482 ) , Abnova ( CC2D1A , H00054862-B01P ) and a customized rabbit polyclonal peptide antibody directed against residues 577–592 ( GenScript , [10] ) Cells were seeded on coverslips in 24 well dishes and cultivated overnight . Cells were then washed in dPBS and fixed for 10 min in cold 4% PFA . After washing and permeabilisation for 10 min in 0 . 3% Triton X-100 cells were incubated in blocking buffer ( 10% NGS , 0 . 1% Tween-20 , 0 . 3 M Glycine in dPBS ) for at least 1 hour at room temperature . Cells were then incubated with primary antibodies ( CC2D1A/mLGD2 ( this study ) or PGK1 ( Gene Tex , GTX107614 ) and CHMP4B ( Santa Cruz , C12 ) ) in staining solution ( 1 . 5% NGS , 0 . 1% Tween-20 in dPBS ) overnight at 4°C , followed by 1 hour incubation with PLA probes . The PLA Probe for the CC2D1A/mLGD2 antibody was generated using the Duolink In Situ Probemaker Minus Kit ( Olink Bioscience ) and anti-guinea pig IgG ( Jackson Immuno Research ) according to manufacturers`instructions and for the CHMP4B antibody the Duolink In Situ PLA probe anti-rabbit was used . The next steps were performed according to the Duolink In Situ protocol provided by Olink Bioscience . For detection the Duolink In Situ detection reagents red were used . Statistical analysis was performed using Student`s t-test , unpaired , two-tailed . Cells were grown to confluency in 96 well dishes on Aclar films , fixed for 30 min in fixing solution ( 2 . 0% glutaraldehyde , 0 . 2% saturated picric acid in 0 . 1 M cacodylate buffer ) washed in 0 . 1 M cacodylate buffer and postfixed in 1% osmium tetroxide in 0 . 1 M cacodylate buffer for 30 min . The specimens were dehydrated in EtOH and embedded in Epon using ethanol as an intermediate solvent . Thin sections were contrasted for 5 min in 2% uranyl acetate and 4 min in Reynolds lead citrate and observed under an EM 902 ( Zeiss ) microscope at 80 KV . For the quantification of the perimeter of endosomal and lysosomal areas , images of whole cells were acquired , as well as higher magnifications of all parts which contained endosomal compartments . ImageJ was used to trace the membrane of endosomes and lysosomes and the perimeter was measured . At least four individual experiments were performed and at least 35 cells for each genotype were analysed . Statistical analysis was performed using Student`s t-test , unpaired , two-tailed . The human CC2D1A and CC2D1B constructs were generated by PCR using synthesized CC2D1B ( Ensembl transcript ENST00000284376 , Genscript ) or CC2D1A ( Clone ID: 6585236 , Source BioScience ) as templates . Amplified sequences were cloned into plgdPattB using NotI and KpnI or XhoI restriction sites [22] . Primer sequences are available upon request . All constructs were sequenced prior to injection . D . melanogaster lines: lgdd7 FRT40A [65] and shrub4-1 FRTG13 [27] . Antibody staining was performed according to standard protocols . Anti-wingless antibody was purchased from DSHB ( 4D4 ) and nuclei were stained with Hoechst 33258 dye . Images were obtained with an Apotome microscope ( Axio Imager Z1m , Zeiss ) .
|
The proteins of the Lgd/CC2D1 family are conserved in all multicellular animals . The Drosophila melanogaster ortholog Lgd is involved in the regulation of signalling receptor degradation via the endosomal pathway . Loss of lgd function causes ectopic ligand-independent activation of the Notch signalling pathway due to a defect in the endosomal pathway . For the mammalian proteins no endosomal function has been defined so far . Here , we asked whether the function of Lgd is conserved in mammals with the focus on the question whether its orthologs are also involved in the endosomal pathway and regulation of Notch pathway activity . Therefore , we generated and characterised Cc2d1a and Cc2d1b conditional knockout mice . We found that the loss of Cc2d1b does not lead to an obvious phenotype , while the known lethality of Cc2d1a deficient newborns is nervous system dependent . In experiments with MEFs isolated from knockout animals we provide evidence that both CC2D1 proteins are involved in the function of the ESCRT-III complex in a similar manner as Lgd in D . melanogaster . Moreover , we found that the loss of one CC2D1 protein is not sufficient to cause ectopic activation of Notch signalling .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
The Mammalian Orthologs of Drosophila Lgd, CC2D1A and CC2D1B, Function in the Endocytic Pathway, but Their Individual Loss of Function Does Not Affect Notch Signalling
|
To understand the interplay of residual structures and conformational fluctuations in the interaction of intrinsically disordered proteins ( IDPs ) , we first combined implicit solvent and replica exchange sampling to calculate atomistic disordered ensembles of the nuclear co-activator binding domain ( NCBD ) of transcription coactivator CBP and the activation domain of the p160 steroid receptor coactivator ACTR . The calculated ensembles are in quantitative agreement with NMR-derived residue helicity and recapitulate the experimental observation that , while free ACTR largely lacks residual secondary structures , free NCBD is a molten globule with a helical content similar to that in the folded complex . Detailed conformational analysis reveals that free NCBD has an inherent ability to substantially sample all the helix configurations that have been previously observed either unbound or in complexes . Intriguingly , further high-temperature unbinding and unfolding simulations in implicit and explicit solvents emphasize the importance of conformational fluctuations in synergistic folding of NCBD with ACTR . A balance between preformed elements and conformational fluctuations appears necessary to allow NCBD to interact with different targets and fold into alternative conformations . Together with previous topology-based modeling and existing experimental data , the current simulations strongly support an “extended conformational selection” synergistic folding mechanism that involves a key intermediate state stabilized by interaction between the C-terminal helices of NCBD and ACTR . In addition , the atomistic simulations reveal the role of long-range as well as short-range electrostatic interactions in cooperating with readily fluctuating residual structures , which might enhance the encounter rate and promote efficient folding upon encounter for facile binding and folding interactions of IDPs . Thus , the current study not only provides a consistent mechanistic understanding of the NCBD/ACTR interaction , but also helps establish a multi-scale molecular modeling framework for understanding the structure , interaction , and regulation of IDPs in general .
It is now widely recognized that many functional proteins lack stable tertiary structures under physiological conditions [1]–[5] . Importantly , such intrinsically disordered proteins ( IDPs ) are highly prevalent in proteomes [6] , play crucial roles in cellular areas such as signaling and regulation [7] , [8] , and are often associated with human diseases such as cancers [9]–[11] . The concept that intrinsic disorder can confer functional advantages has been discussed extensively [12]–[16] . For example , the disordered nature of IDPs could offer several unique benefits for signaling and regulation , including high specificity/low affinity binding , inducibility by posttranslational modifications , and structural plasticity for binding multiple partners . The last property appears to be particularly advantageous , and could support one-to-many and many-to-one signaling [16] , [17] . Nonetheless , the physical basis of these proposed phenomena remains largely elusive . Specifically , how IDP recognition and regulation are supported by the interplay of residual structures , conformational fluctuations and other physical properties as encoded in the peptide sequence is poorly understood . The current limit in mechanistic understanding of how intrinsic disorder supports function might be attributed to two key challenges in characterizing IDPs . These challenges are broadly shared by mechanistic studies of protein folding , misfolding , and aggregation in general [18]–[21] . The first one is related to the difficulty in deriving detailed structural information of the disordered unbound states [22]–[25] . In general , only ensemble-averaged properties can be measured for disordered proteins except with single-molecule techniques ( which have their own limitations in spatial resolution , labeling need , and protein size [26]–[28] ) . Recovering the underlying structural heterogeneity using averaged properties is a severely underdetermined problem [29]–[33] . It is generally not feasible to construct a unique disordered structure ensemble that is consistent with the available data . This fundamental limitation leads to significant ambiguity in the current knowledge of the conformational nature of unbound IDPs . The second challenge is to further clarify the functional roles of any putative conformational sub-states or other properties of an IDP in its recognition and regulation ( i . e . , “function” ) . In particular , whereas some IDPs remain disordered in complexes [34] , [35] , many fold into stable structures upon binding to specific targets [36] . The roles of intrinsic disorder vs . residual structures in such coupled binding and folding interactions have been under much debate [36] . On one hand , residual structures have been observed frequently in unbound IDPs , and intriguingly , such residual structures often resemble those in the folded complexes [37]–[41] . These observations have led to an attractive hypothesis that preformed structural elements might provide initial binding sites to facilitate efficient recognition ( i . e . , conformational selection-like mechanisms ) [12] , [37] . On the other hand , evidence has accumulated in recent years , from computation as well as experimentation , to support a central role of nonspecific binding and emphasize the importance of disordered nature itself in promoting facile IDP recognition [41]–[51] . In fact , all published studies that extend beyond examining the unbound states alone have suggested induced folding-like mechanisms , at least at the baseline level . Precisely how the disordered nature contributes to binding , however , is less clear . One proposal is that nonspecific binding of unstructured and presumably more extended conformations can increase the capture radii to enhance the binding kinetics [52] , [53]; however , such “fly-casting” effects is small with a theoretical maximum of ∼1 . 6-fold acceleration . Recent studies have shown that unbound IDPs tend to be much more compact than previously assumed [54]–[57] , further reducing the proposed fly-casting affects . In addition , the rate-enhancing affect due to increased size is likely offset by slower diffusion [58] . Alternatively , the unbound state of IDPs is presumed heterogeneous and strongly fluctuating . More specifically , conformational sub-states in the unbound IDPs should be marginally stable and separated by small free energy barriers ( e . g . , a few kcal/mol or less ) . These conformational fluctuations could contribute to efficient IDP recognition by allowing the peptide to fold rapidly upon ( nonspecific ) binding [50] , [58] , which is required for achieving the diffusion-controlled maximum binding rate ( otherwise folding becomes rate-limiting ) [59] . It should be noted that cellular events frequently modify the folding of IDPs to modulate their activities , such as through phosphorylations or by binding of other proteins [60] . Therefore , in contrast to globular proteins where folding often serves only to achieve the native structures , folding and unfolding appears to be direct and inherent aspects of IDP function . This underpins the importance and biological relevance of obtaining a mechanistic understanding of binding-induced folding of IDPs beyond a subject of theoretical curiosity . The challenge in detailed characterization of IDPs represents a unique opportunity for molecular modeling to make critical contributions [5] . In particular , atomistic simulations could provide the ultimate level of detail necessary for understanding the structure and interaction of IDPs . At the same time , the dynamic and heterogeneous nature of IDPs also pushes the limits of both the force field accuracy and conformational sampling capability . So-called implicit solvent is arguably an optimal choice for de novo simulations of IDPs because of its necessary balance of accuracy and speed [61]–[64] . The basic idea of implicit solvent is to capture the mean influence of water by direct estimation of the solvation free energy , therefore reducing the system size about 10-fold . Important advances have been made to greatly improve the efficiency and achievable accuracy of implicit solvent , such as via the popular generalized Born ( GB ) theory [64] . With reduced system size , implicit solvent is also particularly suitable for replica exchange ( REX ) simulations [65]–[67] , an enhanced sampling technique that has proven highly effective in sampling protein conformational equilibria [68] . Importantly , improved efficiency with implicit solvent also allows careful optimization to suppress certain systematic biases that have plagued explicit solvent approaches [69] , [70] . For example , we have previously optimized the generalized Born with smooth switching ( GBSW ) model [71] , [72] together with the underlying CHARMM22/CMAP protein force field [73]–[76] . The resulting GBSW protein force field not only recapitulates the structures and stabilities of helical and β-hairpin model peptides with a wide range of stabilities [77] , [78] , but also allows calculation of the conformational equilibria of small proteins under stabilizing and destabilizing conditions [79]–[81] . Although inherent and methodological limitations remain in implicit solvent [82] , initial applications of implicit solvent to modeling small IDPs have been reasonably successful [41] , [46] , [55] , [83]–[86] , substantiating the notion that it is a viable approach for atomistic simulations of IDPs . The current work focuses on the nuclear-receptor co-activator binding domain ( NCBD ) of the transcription coactivator CREB-binding protein ( CBP ) and its interaction with the p160 steroid receptor co-activator ACTR . CBP and its paralogue p300 are general transcriptional coactivators that play critical roles in transcriptional regulation and participate in cell cycle control , differentiation , transformation , and apoptosis [87] , [88] . The NCBD domain ( residues 2059–2117 in mouse CBP ) is also known as interferon regulatory factor ( IRF ) binding domain ( iBID ) or the SRC1 interaction domain ( SID ) . It mediates the interaction of CBP with a number of important proteins , including steroid receptor coactivators , p53 and IRFs [2] , [89] . The interaction of CBP with p160 coactivators in particular is important for recruitment of CBP/p300 to transmit the hormonal signal to the transcription machinery [90] . Besides the biological and medical significance , the NCBD/ACTR interaction also offers unique opportunities for understanding the molecular principles of IDP recognition . Both NCBD and the activation domain of ACTR that it interacts with ( residues 1018–1088 in human ACTR; hereafter referred to as ACTR ) are IDPs . Their interaction is an example of the “synergistic folding” mechanism [91] ( the other known example also involves NCBD , but with the p53 transactivation domain , TAD [92] ) . In addition , four folded structures of NCBD have been solved in complex with various protein targets besides ACTR [91]–[94] . In these complexes , NCBD adopts two distinct tertiary folds that involve three similar helices , represented by the NCBD/ACTR and NCBD/IRF3 complexes ( see Figure 1 ) . Therefore , NCBD represents one of the few experimentally validated examples of structural plasticity , which is believed to be a key functional advantage of intrinsic disorder [16] . Interestingly , although free ACTR is largely devoid of residual structures , free NCBD contains one the highest levels of residual structures with folded-like helical content and molten globule characteristics [95] , [96] . In addition , even though nuclear magnetic resonance ( NMR ) relaxation analysis has established that free NCBD is highly dynamic on picosecond ( ps ) to nanosecond ( ns ) timescales [96] , it appears to have a strong tendency to adopt marginally stable tertiary folds , allowing two NMR structures of the unbound state determined to date [40] , [89] . These structures are presumably obtained by stabilizing various conformational sub-states under specific solution conditions . Particularly intriguing is that the latest NMR structure of free NCBD turns out to be similar to the folded conformation observed when bound to ACTR [40] . Although such pre-existence of folded-like conformations should be considered only as a necessary but insufficient condition for conformational selection-like mechanisms , the unusually high level of residual structures of NCBD strongly suggests a functional role of pre-folding in its coupled binding and folding interactions . In this work , we first exploit implicit solvent-based atomistic simulations and REX enhanced sampling to characterize the conformational properties of free NCBD and ACTR . The roles of preformed structures vs . conformational fluctuation in the NCBD/ACTR interaction are then directly probed using high-temperature unfolding and unbinding simulations in both implicit and explicit solvents . Combined with our recent coarse-grained simulations and existing experimental data , we aim to obtain a detailed mechanistic picture of how residual structures , conformational fluctuations , and electrostatic interactions contribute to efficient synergistic folding of NCBD and ACTR .
De novo calculation of the disordered ensembles for IDPs is challenging [5] , especially for NCBD that is both of moderate size and apparently with a complex , solution condition-sensitive conformational equilibrium . Our previous works have suggested that implicit solvent coupled with REX enhanced sampling could generate reasonably accurate disordered ensembles for small IDPs , including a 28-residue segment of the kinase inducible domain ( KID ) of transcription factor CREB [55] . In Figure S1 , we first test the convergence of the calculated disordered ensembles by examining the dependence of residue helicity on REX simulation time and by comparing results from independent simulations initiated from dramatically different conformations ( folding vs . control; see Methods ) . The sequences of both domains are provided in Methods . Free ACTR appears to be highly disordered with marginal residual helicity . The calculated residual helicity profiles from the control and folding runs converge to similar ones ( data not shown ) . For NCBD , while the time evolution of the calculated residual helicity appears to stabilize over the course of 100 ns in either the control or folding REX simulation , the final profiles from these two independent calculations differ substantially , suggesting that the actual convergence is rather limited . Nonetheless , both the folding and control simulations clearly suggest significant residual helicity in all three helical segments that become stably folded upon binding to various specific targets . Detailed analysis of the conformational ensemble ( see below ) demonstrates that free NCBD is compact and contains substantial tertiary contacts . These conformational properties of NCBD , coupled with the larger size , contribute to the difficulty of achieving better convergence using the REX/GB protocol . In addition , the current surface area-based treatment of nonpolar solvation can over-stabilize non-specific collapsed states [82] , [97] . This problem further limits the ability to sufficiently sample accessible tertiary organizations of free NCBD and their inter-conversions , which is required for achieving good convergence . Given the limited convergence achieved in the REX simulations of free NCBD and apparent difficulties in substantially improving the level of convergence , we focus on semi-quantitative or qualitative analysis of the conformational properties of NCBD . That is , although significant conformational sub-states sampled by REX may be genuine , the relative stability ( population ) is not likely to be reliable . Considering that NCBD is experimentally known to be highly helical , the folding simulations ( initiated with a fully extended conformation ) should take longer to converge , and the disordered ensemble calculated from the control simulation is likely more realistic . Therefore , all the subsequent analysis is based on the ensemble of conformations sampled during the last 60 ns of the 100 ns control REX simulation . In Figure 2 , we compare the residue helicity of NCBD and ACTR in the free and bound states . The results appear to be fully consistent with the previous NMR secondary chemical shift analysis ( Figure 2 of Ref . [96] ) , showing that all three NCBD helices are largely formed in the unbound state and ACTR is largely free of residual helices . Interestingly , the poly-Q segment of NCBD ( residues 2082–2086 ) , although disordered in the NCBD/ACTR complex , is largely helical in the unbound state and extends Cα2 . This is fully consistent in the NMR chemical shift analysis [96] . Recent sequence correlation analysis has revealed a link between sequence order and binding promiscuity [98] , [99] . One might expect that the length of the poly-Q stretch might affect conformational flexibility , and furthermore , the ability to interact with diverse targets . We also have analyzed the ensemble distribution of the radius of gyration of free NCBD . The results , shown in Figure S2A , confirm that free NCBD is highly compact . Despite a clear lack of convergence , the control and folding simulations appear to sample a set of conformation sub-states with similar characteristic sizes . Direct comparison of the calculated size profiles to one derived from a recent small-angle X-ray scattering ( SAXS ) study [40] is complicated by the different constructs used and uncertainty in proper inclusion of the solvation shell for a heterogeneous ensemble . Nonetheless , one can estimate that including the disordered N- and C-terminal tails ( 13 residues total ) truncated in the current simulations would increase the radius by 2–3 Å , and that the solvation shell may add another 2–3 Å ( estimated by comparing results from HydroPro [100] and CHARMM ) . These corrections together bring the calculated radius of the gyration profile close to the SAXS-derived profile that centers around 15 . 2 Å under “native-like” conditions [40] . Apparent agreement between NMR and SAXS on these ensemble-averaged properties is not sufficient to validate the reliability of the simulations , but it suggests that the simulated ensemble may offer a qualitative or even semi-quantitative characterization of the conformational properties of free NCBD . Because all three NCBD helices are largely formed in the unbound state , the conformational fluctuation of free NCBD mainly involves tertiary packing of these helices . For example , as shown in Figure 1 , when aligned using the central helix Cα2 , the two representative folded conformations of NCBD differ mainly in the orientation of Cα1 and slightly less so in that of Cα3 . Therefore , all conformations of the calculated ensemble first re-oriented by aligning Cα2 ( to the −z axis ) before the orientations of Cα1 and Cα3 were calculated . Note such analysis also provides an effective description of the tertiary packing even when one or more of the three NCBD segments are not in helical states . The results , shown in Figure 3 , illustrate that NCBD is strongly fluctuating and samples a large number of helix configurations , as expected for a molten globule . Intriguingly , free NCBD appears to substantially sample all three distinct conformations that have been observed experimentally so far , either in complexes or in isolation . These folds are represented by PDB structures 1kbh , 1zoq , and 1jjs , respectively . The Cα1 orientation of 1kbh and Cα3 orientation of 1jjs appear to be least sampled . Nonetheless , conformational sub-states exist with similar orientations , as marked by arrows in panels c ) and d ) of Figures 3 . Specifically , for 1kbh-like Cα1 orientation , the adjacent sub-state contains more parallel ( with smaller helix cross angles ) , and thus tighter , packing of Cα1 with Cα2 , but with a helix interface similar to that of 1kbh . Further structural analysis ( see the following paragraph ) suggests that such tighter packing is likely a result of helix formation in the poly-Q segment ( e . g . , see Figure 2 ) , which shortens the Cα1-Cα2 loop and promotes tighter packing . Clustering analysis was performed to further analyze the structural properties of the major conformational sub-states of free NCBD . The average structures of the six most populated clusters identified using K-means clustering with a 3 . 0 Å radius are shown in Figure 4 . Helix configurations for all members of these clusters are shown in Figure S3 . Interestingly , even though one of the clusters ( Figure 4D ) is similar to the fold observed in 1kbh , most clusters are different from either 1zoq or 1kbh on the whole domain level , as suggested by the large RMSD values . Therefore , even though both individual Cα1-Cα2 and Cα2-Cα3 helix pairs sample all three distinct PDB folds , these folded-like configurations of individual helix pair generally do not occur at the same time . Notably , the folded conformations of NCBD in 1kbh and 1zoq have relatively similar Cα2-Cα3 helix packing ( see Figure 1C ) . The packing of Cα2 and Cα3 also appears to be more restricted in free NCBD compared to that of Cα1 and Cα2 ( e . g . , as indicated by a larger “inhibited” red area in Figure 3D compared to Figure 3C ) . NCBD has a strong inherent propensity to adopt Cα2-Cα3 configurations analogous to those in 1kbh and 1zoq . Such persistent folded-like conformations of free NCBD could contribute to recruitment of specific targets such as ACTR and IRF3 , allowing NCBD to adopt different final structures by docking the more flexibly linked Cα1 into different positions . Another interesting observation is that the poly-Q segment appears to be capable of readily switching between helical and coil states . Such conformational fluctuations could allow NCBD to adapt to different substrates , extending the Cα2 helix when bound with IRF3 but becoming more disordered when in complex with ACTR ( see Figure 1 ) . Although the REX simulations provide intriguing insights into the possible residual structures of free NCBD , how these conformational properties contribute to synergistic folding of NCBD with ACTR is not obvious based on these equilibrium simulations alone . For this , one could calculate the coupled binding and folding free energy surfaces [41] , [51] or transition paths [46] to more directly clarify the recognition mechanism and probe the roles of residual structures vs . pre-folding in specific finding . However , given the moderate size and relatively complex topology , such calculations can be extremely demanding using an atomistic physics-based force field for the NCBD/ACTR complex . Instead , temperature-induced unfolding and unbinding simulations may be used to effectively infer the molecular processes of coupled binding and folding . A key assumption is that binding/folding is largely a reverse of unbinding/unfolding . An important concern is that the transition states or the most probable transition paths might depend on temperature [101] . Nonetheless , high-temperature unfolding simulations have so far proven quite successful for studying folding and interaction of many proteins , including IDPs [44] , [102]–[104] . A 100 ns equilibrium simulation of the complex was first performed at 300 K , which confirms that the native fold ( model 1 of PDB:1kbh ) is very stable in the GBSW/MS2 implicit solvent ( see Figure S4 ) . Subsequent pilot simulations suggest 475 K to be optimal for simulating unbinding and unfolding of the NCBD/ACTR complex in GBSW/MS2 ( e . g . , see Figure S5 ) . In Figure 5 , we compare the time evolution of various fractions of native contacts computed from 50 independent unfolding simulations at 475 K . The fraction of native intermolecular interactions ( Qinter ) is used to describe binding , and the fraction of native tertiary intramolecular interactions ( QNCBD ) is used for folding of NCBD . As shown in Figure S6 , ACTR is completely devoid of any inter-helix tertiary contacts in the NCBD/ACTR complex . Because ACTR is largely free of residual structures in the unbound state , the overall helicity ( αACTR ) is used to effectively monitor ( binding-induced ) folding of ACTR . On the baseline level , all unfolding and unbinding kinetics appear to be reasonably well represented by single exponential functions . The fitted kinetic data is summarized in Table 1 . The secondary ( helix ) unfolding of NCBD is predicted to be the slowest process ( αNCBD; green traces in Figure 5 ) , which is expected given the high level of residual structures in unbound NCBD; however , both the ACTR ( helix ) and NCBD tertiary unfolding appear to be significantly faster than unbinding . This result suggests that binding occurs prior to the folding of both ACTR and NCBD; that is , both ACTR and NCBD follow induced-folding-like mechanisms on the baseline level in the GBSW/MS2 implicit solvent . Considering the apparent tendency of NCBD to pre-fold ( see above ) , this result is somewhat surprising , but it highlights the importance of conformational fluctuations and nonspecific binding in specific recognition of IDPs , even for IDPs with significant residual structures like NCBD . Significant heterogeneity is apparent in the unfolding/unbinding pathways of NCBD/ACTR and is partially reflected in substantial ruggedness that remains in the curves shown in Figure 5 ( e . g . , compared with a previous explicit solvent unfolding simulation of the p53-MDM2 complex , where 10 10-ns simulations at 498 K were sufficient to yield much smoother curves [103] ) . The complex fully disassociates within 10 ns in only 6 out of the 50 independent runs . In examining the unbinding/unfolding characteristics at a lower temperature of 450 K ( see Figure S7 ) , we found the heterogeneity of unfolding/unbinding pathways to be even more evident . In addition , the complex appears trapped in some intermediate states and does not fully unfold/unbind even after 20 ns . Nonetheless , unfolding of either ACTR or NCBD appears to lag behind unbinding , which is consistent with the induced-folding baseline mechanisms predicted at 475 K . Indications are that binding-induced folding of NCBD and ACTR is not simply 2-state-like . For example , decay of QNCBD and αACTR appears to pause at ∼2 ns ( red and blue traces in Figure 5 ) , which could suggest a common intermediate state where ACTR and NCBD are partially bound and folded . The decay curves are too noisy ( partially due to underlying heterogeneity ) for reliable kinetic fitting using double exponential functions . Therefore , we constructed ( pseudo ) unbinding and unfolding free energy surfaces based on statistics collected from the first 5 ns of the unfolding simulations . Note that the system is not at equilibrium during this time frame , so the resulting free energy profiles are not equilibrated ( and thus strongly dependent on initial conditions ) . Nonetheless , the profiles provide qualitative approximations of the true free energy surfaces [105] . As shown in Figure 6A , an intermediate state is evident at Qinter∼0 . 25 and QNCBD∼0 . 15 . Interestingly , a similar key intermediate state also has been predicted in our recent topology-based modeling of the NCBD/ACTR complex [50] . A strong resemblance between the free energy surface is shown in Figure 6A and the result derived from topology-based modeling ( Figure 5A of reference [50] ) . Both the atomistic simulations ( see further analysis detailed in the following paragraph ) and topology-based modeling predict that the intermediate state mainly involves the C-terminal segments of NCBD and ACTR . Such a prediction appears highly consistent with a recent H/D exchange mass spectrometry ( H/D-MS ) study [106] , where peptide segments within the C-terminal regions of both NCBD and ACTR were found to have much larger protection factors compared with those mapped into other folded regions of the complex . In Figure 7 , we further examined the binding kinetics of individual NCBD and ACTR helices . The kinetic data derived from fitting to single exponential functions is summarized in Table 1 . The analysis shows that Aα3 and Cα3 unbind with the largest half times , τ = 2 . 93 ns and 2 . 20 ns , respectively , which are greater than that of the overall intermolecular interaction formation ( τ = 1 . 61 ns ) . This result indicates that binding is mainly initiated by the C-terminal helices . In contrast , the first helices of NCBD and ACTR unbind much faster then the second and third helices . In fact , unbinding of Aα1 and Cα1 occurs even faster than folding of either NCBD or ACTR ( as described by QNCBD and αACTR , see Table 1 ) . These kinetic rates are consistent with a multi-stage synergistic folding process , where NCBD and ACTR first bind rapidly through the C-terminal segments , forming intermediates that are mainly stabilized by native-like interactions between α2 and α3 helices . This first step appears to be highly cooperative ( e . g . , see Figure 6A ) , although indications are that both induced folding and conformational selection might contribute [50] . Interestingly , the transition between the intermediate and bound states appears largely conformational selection-like where NCBD and ACTR folding precedes Aα1 and Cα1 binding . Formation of the partially folded core appears to facilitate the rest of NCBD to fold into native-like conformations , allowing Cα1 and Aα1 to rapidly form native intermolecular interactions en route to the fully folded bound state . Taken together , even though the synergistic folding of NCBD and ACTR follows an induced folding-like baseline mechanism ( where binding precedes folding on the overall level ) , detailed analysis reveals multiple stages of induced folding and conformational selection . Such a mechanism closely resembles an “extended conformational selection” recently proposed by Csermely et al . [107] , [108] and is remarkably consistent with our recent topology-based modeling of the NCBD/ACTR complex [50] . One of the most notable features of the NCBD/ACTR complex is a buried salt-bridge between NCBD R2105 and ACTR D1068 [91] ( see Figure 1A ) , which is also conserved in the interaction of NCBD with p53 TAD [92] . Interestingly , this buried salt-bridge is part of a local network of salt-bridges that could form between multiple complementary charges , including R2105 and K2108 of NCBD and D1060 , E1065 , and D1068 of ACTR ( see Figure 1A ) . This network of native and non-native salt-bridges appears to play a significant role in stabilizing the putative intermediate state , either thermodynamically or kinetically . Although most individual salt-bridges frequently break and reform during individual unfolding simulations ( see Figure S8 ) , on average they largely persist throughout the 10 ns unfolding simulations at 475 K and hinder the transition from the partially bound intermediates to fully disassociated ones ( see Figure 8 ) . Out of the 50 unfolding simulations at 475 K , the complexes fully dissociate only by the end of 10 ns simulations in six cases . The native salt-bridges , between NCBD R2105 and ACTR D1068 and D1060 , are the most protected . As shown in Figure 8 , they are the most preserved and remain formed over 80% of the time throughout the simulations ( blue and black traces in Figure 8A ) . NCBD K2108 is adjacent to R2015 and close enough to interact with ACTR D1068 and D1060 , but these salt-bridges are more solvent-exposed and thus slightly less preserved during high-temperature simulations . The side chain of ACTR E1065 is positioned away from NCBD in the native structure . Partial unfolding of Aα2 allows E1065 to rotate and participate in the salt-bridge network with 10–30% probability by the end of the 10 ns simulation at 475 K ( purple and red traces in Figure 8A ) . The conformational heterogeneity of the intermediate state does not permit reliable free energy calculations to quantify the contribution of salt-bridge interactions to stability . Nonetheless , previous mutagenesis studies have suggested that the buried salt-bridge between NCBD R2105 and ACTR D1068 contributes minimally to binding affinity [95] . The salt-bridge network likely could not significantly stabilize the intermediate state thermodynamically , either , which raises a concern that the observed persistence of the local salt-bridge network is artificial , such as due to over-stabilization of charge-charge interactions in the GBSW/MS2 implicit solvent . To address this concern , we first examine the potential of mean forces ( PMFs ) between Arg and Asp side chain analogs in TIP3P and GBSW/MS2 . The results , summarized in Figure 8 , show that GBSW/MS2 actually slightly under-stabilizes the Arg-Asp interaction compared with TIP3P , either in a constrained head-to-head configuration ( which was used in the force field optimization [72] ) or when fully unconstrained . In particular , configurationally unconstrained Arg-Asp interaction is unstable in GBSW/MS2 ( Figure 9B ) . Therefore , the observed stabilization effects of salt-bridges on the intermediates are likely of a kinetic nature . Such kinetic stabilization arises from substantial desolvation barriers in disassociation of salt-bridges , particularly in partially folded protein environments where the side chain configurations are restricted ( e . g . , see Figure 9A ) . With a concentrated local network of salt-bridges , very large desovaltion barriers can be expected for complete dissociation of NCBD and ACTR , which explains why only a small fraction of the high-temperature simulations ( 6 out of 50 ) successfully reached the fully unbound state in 10 ns . To further confirm that the observed salt-bridge network is not an artifact of implicit solvent , a set of 10 unfolding simulations was performed in TIP3P explicit solvent at 500 K . Most simulations were terminated between 3 to 4 ns when the complex size exceeded the periodic box dimensions . The lengths of these simulations are insufficient to capture degrees of unfolding and unbinding similar to implicit solvent simulations , and the number of trials is insufficient to obtain smooth curves for kinetic fitting . Nonetheless , visual inspection of simulation trajectories as well as examination of the evolution of various contact fractions support an unbinding and unfolding mechanism that is consistent with the one derived from implicit solvent simulations ( see Figure S9 ) . The same set of native and non-native interactions , particularly the buried one between NCBD R2105 and ACTR D1068 ( blue trace in Figure S9B ) , persist and appear to stabilize the partially unbound and unfolded intermediates . Note that the helical secondary structures are substantially over-stabilized in these explicit solvent simulations ( e . g . , see the blue trace in Figure S9A ) . This is a known artifact of the current version CHARMM22/CMAP explicit solvent force field [78] , [109] , [110] . A control simulation of the double-Leu mutant complex , NCBD:R2105L/ACTR: D1068L , at 300 K suggests that the native fold remains stable in the GBSW/MS2 implicit solvent ( data not shown ) . A set of 50 unfolding simulations was carried out at 450 K to further investigate the role of the buried salt-bridge in synergistic folding . The heterogeneity of the unfolding/unbinding pathway observed in the wild-type complex ( e . g . , see Figure 5 ) is even more pronounced without the buried salt-bridge . All averaged time traces of contact fractions remain very noisy ( e . g . , see Figure S10 ) . Most traces cannot be satisfactorily fitted to either single or double exponential functions , preventing quantitative analysis of unfolding and unbinding kinetics . Nonetheless , the pseudo binding and folding free energy surface computed from the first 5 ns of the unfolding trajectories appears to resemble that from simulations of the wild-type complex ( see Figure 6 ) . In particular , a similar intermediate state exists at Qinter∼0 . 2 and QNCBD∼0 . 15; however , the small free energy barrier separating the intermediate and fully unbound states in Figure 6A is largely absent in Figure 6B . Removal of NCBD:R2105L largely disrupts the local salt-bridge network . The intermediate state appears to have much shorter resident times , and can quickly fluctuate to the fully unbound state . Importantly , examination of the evolution of intermolecular contact factions of individual NCBD and ACTR helices , shown in Figure S10 , supports that the mutant complex largely follows a similar , albeit more heterogeneous , unbinding and unfolding mechanism , with the N-terminal α1 helices disassociated first ( black traces in Figure S10B–C ) . These results suggest the local salt-bridge network does not appear to fundamentally modulate the recognition mechanism . Instead , it mainly augments a productive synergistic folding mechanism inherent in ( the topology of ) the NCBD/ACTR complex , by transiently stabilizing a key on-pathway intermediate state to facilitate complete folding en route to the specific complex .
With one of the highest levels of residual structures , NCBD is an intriguing model system for understanding the roles of residual structure vs . conformational fluctuations in coupled binding and folding of IDPs . We have combined equilibrium and non-equilibrium simulations using physics-based , atomistic protein force fields to characterize the conformational properties of unbound NCBD and ACTR and to understand how these properties facilitate efficient synergistic folding of these two IDPs . The calculation recapitulates that free NCBD has folded-like helical content , is strongly fluctuating , and samples a wide range of tertiary configurations , which is consistent with the previous notion that free NCBD is a molten globule [96] . Intriguingly , the calculated disordered ensemble of NCBD contains significant populations with helical packings that are highly similar to all those previously observed experimentally in isolation and in complex with various targets . Observations of such pre-folded conformations , especially for IDPs with significant residual structures like NCBD , could be considered strong evidence for conformational selection-like mechanisms , where such preformed structural elements provide initial binding sites . Direct examination of the unfolding and unbinding pathways in high-temperature simulations , however , shows that both ACTR and NCBD tend to unfold first before unbinding , suggesting an induced folding-like baseline mechanism for their synergistic folding . This seemingly surprising result appears to be consistent with the observation that , although individual Cα1/Cα2 and Cα2/Cα3 helical pair samples folded-like packing with substantial probability , these configurations rarely occur simultaneously . Therefore , population of folded-like tertiary conformations on the whole domain level is insufficient to support conformational selection-like mechanisms on the baseline level . Further analysis reveals an on-pathway intermediate state that mainly involves the C-terminal helices of ACTR and NCBD , which also has been predicted by a recent coarse-grained simulation study using topology-based models [50] . Importantly , existence of such a major intermediate state also appears to be consistent with a recent H/D-MS experiments showing that peptide segments within the C-terminal regions of NCBD and ACTR have much larger protection factors compared with those mapped into other regions of the complex [106] . Our kinetic analysis suggests that , once the initial mini folding core is formed , the N-terminal helix of NCBD folds rapidly ( Table 1 ) , allowing subsequent facile binding and folding the ACTR N-terminal helix en route to the final specific complex . Therefore , although the baseline mechanism is induced folding-like , conformational selection actually occurs at local levels . Together with our recent topology-based modeling study [50] , the atomistic simulations strongly support the prediction that synergistic folding of NCBD and ACTR follows the “extended conformational selection” mechanism [107] . Our topology-based modeling of the NCBD/ACTR interaction [50] has revealed a separate , albeit less prevalent , pathway where binding is initiated by the N-terminal α1 helices . These mechanistic insights on synergistic folding of NCBD and ACTR , derived from the atomistic and coarse-grained simulations , are summarized in Figure 10 . An intriguing interplay appears to exist among residual structures , conformational fluctuations , and electrostatic interactions to facilitate the rate-limiting step of forming the partially folded intermediates . The NCBD Cα2/Cα3 helix-turn-helix motif appear to be conformationally more restricted ( Figure 2D ) , whereas the C-terminus of Cα3 retains the least amount of helical content and is considerably more heterogeneous ( Figure S2B ) . Both features were also observed in the previous NMR chemical shift and relaxation analysis [96] . Such a balance of residual structures and conformational fluctuations is likely important for the NCBD C-terminal to act as a key initiation point for coupled folding and binding to ACTR and other proteins . Another novel insight provided by the current atomistic simulations is the role of a local network of native and non-native salt-bridges in transiently stabilizing the intermediates . These salt-bridge interactions likely do not contribute substantially to the thermodynamic stability of either the intermediates or the final specific complex [95] , but substantial desolvation barriers involved in breaking up these interactions in a conformationally restricted protein environment ( e . g . , Figure 9A ) can extend the resident time of the intermediates to allow the rest of the complex to fold with higher efficiency . As demonstrated using a dual-transition state kinetic model [59] , efficient folding upon encounter is necessary for achieving facile binding at or near the diffusion-limited basal binding rate , a highly desirable property for signaling and regulatory IDPs that need to constantly evade protein degradation machinery in cell . IDPs are known to be enriched with charges [6] . NCBD and ACTR are no exceptions , with +6 and −8 net charges , respectively ( including the flanking loops that remain disordered in the complex [91] ) . These enriched charges hinder ( independent ) folding and can protest against aggregation . In addition , long-range electrostatic interactions between these large numbers of complementary charges on NCBD and ACTR could dramatically enhance the encounter rate , similar to electrostatic steering , which is known to be important in interactions of globular protein [111] . Furthermore , the complementary pattern of charge , especially within the predicted mini folding core involving the C-termini ( Figure 1 ) , suggests that long-range electrostatic interactions could further promote folding-competent encounter complexes before transiently stabilizing the on-pathway intermediates via formation of short-range salt-bridge network . These effects can enhance the efficiency of folding upon encounter to promote facile recognition . The current study also reveals important limitations in both the protein force field accuracy and sampling capability , especially for modeling IDPs of moderate sizes and with complex residual structures . These limitations underscore the importance of continual development of the protein force field , with increased focus on balancing various competing interactions to allow an accurate description of not only a few ( native ) folds but also the whole conformational equilibrium [82] , [112] . Sampling methodologies clearly need to improve . The standard temperature REX-MD has failed to achieve convergence for the disordered ensemble of NCBD within 100 ns . Besides limited simulation timescale , certain limitations of the implicit solvent protein force field also contributed . In particular , current empirical protein models have been shown to contain a systematic bias to over-stabilize protein-protein interactions [113] , [114] . Furthermore , simple surface area-based estimation of the nonpolar solvation free energy employed in most current implicit solvent models also tends to over-stabilize nonspecific compact protein states [82] . The standard temperature REX-MD clearly has limited ability to sample alternative deeply trapped low energy states with high efficiency . These limitations together have also prevented us from more directly investigating the proposed mechanistic roles of electrostatic interactions using atomistic simulations . Despite these outstanding limitations , the key mechanistic features derived from atomistic physics-based simulations , coarse-grained topology-based modeling , and various biophysical measurements are remarkably consistent , which suggests that an integration of multi-scale modeling and experimentation can provide a viable approach for understanding the functional and control of IDPs .
Only segments of the NCBD and ACTR domains that are structured in the complex are included in the current simulations , which include residues 2066–2112 for NCBD ( in mouse CBP numbering; SALQD LLRTL KSPSS PQQQQ QVLNI LKSNP QLMAA FIKQR2105 TAKYV AN ) and residues 1040–1086 for the ACTR domain ( in human ACTR numbering; E GQSDE RALLD QLHTL LSNTD ATGLE EID1068RA LGIPE LVNQG QALEP K ) . The peptide termini are neutralized using with either acetyl ( Ace ) or amine ( NH2 ) groups . A previously optimized GBSW/MS2 model was used in all implicit solvent simulations unless otherwise noted [72] . This model adopts an effective approximation of the molecular surface for defining the solute-solvent boundary , which is believed to be more physical compared to the van der Waals-like surface used in the original GBSW model [115] , [116] . Importantly , the GBSW/MS2 model has also been carefully optimized to balance solvation and intramolecular interactions and can reasonably capture the competition between α and β secondary structures . Specifically for NCBD/ACTR , the structure of the complex ( PDB: 1kbh [91] ) remains stable in the GBSW/MS2 force field for over 100 ns , but substantially deviates from the native conformation in the original GBSW protein force field ( see Figure S4 ) . REX was used to enhance the sampling of the accessible conformational space of free NCBD and ACTR . For this , the Multiscale Modeling Tools for Structural Biology ( MMTSB ) toolset [117] ( http://www . mmtsb . org ) was used in conjunction with CHARMM [118] , [119] . The basic idea of REX is to simulate multiple non-interacting replicas at different temperatures simultaneously . Periodically , one attempts to exchange the simulation temperatures between pairs of replicas based on a Metropolis criterion derived from the detail balance principle . As such , not only the resulting random walk in the temperature space facilitates the system to cross the energy barriers and exploit the conformational space more efficiently , but proper canonical ensembles are also generated at all temperatures , allowing direct calculation of thermodynamic properties for comparison with experiments . We performed two independent REX simulations for each peptide , initiated from the folded structure extracted from the complex ( control ) and a fully extended conformation ( folding ) , respectively . Comparison of the calculated structure ensembles from these independent control and folding runs with dramatically different initial conditions allows rigorous assessment of the convergence . In each REX simulation , 16 replicas were simulated at temperatures exponentially distributed from 270 to 500 K . SHAKE [120] was applied to fix the lengths of all hydrogen-related bonds , allowing a 2 . 0 fs molecular dynamics ( MD ) time step . Temperature exchanges between neighboring replicas were attempted every 2 ps , and the total length of each REX simulation was 100 ns ( 50 , 000 REX cycles ) . Similar REX/GBSW protocols have proven effective in calculating the disordered structural ensembles for other IDPs ( albeit of smaller sizes than NCBD and ACTR studied in the current work ) [41] , [55] . All analysis was performed based on the conformations sampled during the last 60 ns of the control simulation at 305 K ( where most existing experimental data were acquired ) , unless otherwise noted . The orientations of helical segments ( 1044–1058 , 1063–1071 , 1072–1080 in ACTR; 2067–2076 , 2086–2091 , 2095–2110 in NCBD ) were calculated using the Chothia-Levitt-Richardson algorithm [121] as implemented in CHARMM . The K-means clustering algorithm as implemented in the MMTSB toolset was used to cluster the calculated disordered ensembles based on mutual Cα RMSD distances . Various clustering radii ranging from 1 . 5 to 4 . 5 Å were tested before an optimal radius of 3 . 0 Å was used for the final clustering results presented . All molecular visualizations were generated using the VMD software [122] . The same peptide segments defined above were included the simulations of the complex . The model 1 from the NMR ensemble ( PDB: 1kbh ) was first equilibrated in the GBSW/MS2 implicit solvent using energy minimization and short MD with weak harmonic positional restraints imposed on all backbone heavy atoms . Subsequently , a 160 ns unrestrained simulation was performed at 300 K to examine the structural stability and dynamics of the complex near its native basin . The native structure of the NCBD:R2105L/ACTR:D1068L double-Leu mutant complex was prepared by computational mutagenesis and then equilibrated using a similar protocol as described above . To identify the optimal temperatures for unbinding/unfolding simulations , a series of pilot simulations was performed at temperatures ranging from 350 K to 500 K ( e . g . , see Figure S5 ) . At the optimal temperature , the complex should unfold/unbind within tractable time scales ( e . g . , 10–20 ns ) while retaining important details of the unfolding/unbinding pathways . Once such optimal temperatures were chosen ( 450–475 K for the wild-type and 450 K for the mutant ) , 50 independent high-temperature simulations of 10–20 ns in length were initiated from the equilibrated native structures with different initial velocities . The results presented in this work are averages computed from 50 unfolding simulations unless otherwise noted . For native fraction analysis , a list of native tertiary contacts ( shown in Figure S6 ) was first identified using the equilibrated native structure based on side chain minimal heavy atom distances with a 4 . 2 Å cutoff . The native contacts were then divided into inter-molecular and intra-molecular categories . In analysis of the high-temperature simulation trajectories , a contact was considered formed when the minimal heavy atom distance between two side chains was no greater than 4 . 5 Å . Helicity of various helical segments was calculated based on the hydrogen bonding patterns using the COOR SECS module of CHARMM . Additional high-temperature unfolding and unbinding simulations of the wild-type complex were performed in TIP3P water to examine the unfolding/unbinding pathway and in particular the putative role of the buried salt-bridge between NCBD:R2105 and ACTR:D1068 in ( transiently ) stabilizing the intermediate state ( s ) . For this , the equilibrated NCBD/ACTR complex was placed in a cubic water box with periodic boundary conditions imposed . The final solvated system contains 9176 TIP3P water molecules and the box size is ∼65 Å . Two potassium ions were added to neutralize the total charge . The proteins were described by the CHARMM22/CMAP protein force field [73]–[76] . The particle mesh Ewald method was used for long-range electrostatic interactions [123] , and the van de Waals interactions were smoothly switched off from 12 to 13 Å . Lengths of all hydrogen-related bonds were kept constant with SHAKE [120] , and the MD time step was 2 fs . After 10 ps of NPT equilibration at 300 K , a set of 10 independent NVT productions was carried out at 500 K up to 10 ns until the dimensions of the proteins exceed those of the periodic box . The dynamic time step was reduced to 1 fs in the NVT production simulations for numerical stability . An umbrella sampling protocol [77] was used to compute the PMFs between the side chains of Asp and Arg , either constrained in a head-to-head configuration [77] ( see Figure 9 ) or allowed to freely rotate . In the constrained setup , the side chains were allowed to move only in fixed orientations along the reaction coordinate ( indicated by a dashed line in Figure 9 ) , enforced using the MMFP module in CHARMM . For explicit solvent simulations , solutes were solvated by either ∼710 TIP3P waters in a rectangular box ( for the constrained PMF ) or by ∼1040 TIP3P waters in a truncated octahedral box ( for the unconstrained PMF ) . Periodic boundary conditions were imposed . Non-bonded and other setups are identical to those described above for explicit solvent high-temperature simulations . Harmonic restraint potentials were placed every 0 . 5 Å along the reaction coordinate with a force constant of 5 . 0 kcal/mol/Å2 . For each umbrella-sampling window , the system was first equilibrated for 60 ps , followed by 2 ns ( constrained PMF ) or 4 ns ( unconstrained PMF ) NPT production at 300 K and 1 atm . The final PMFs were calculated using the weighted histogram analysis method ( WHAM ) [124] . The constrained PMF in GBSW/MS2 was computed by direct translation of the side chains along the reaction coordinate , and the unconstrained PMF in GBSW/MS2 was computed in the same umbrella sampling protocol except that implicit solvent was used instead of TIP3P waters . Convergence of the PMFs was examined by comparing results from the first and second halves of the data and was shown to be on the order of 0 . 2 kcal/mol .
|
Intrinsically disordered proteins ( IDPs ) are now widely recognized to play fundamental roles in biology and to be frequently associated with human diseases . Although the potential advantages of intrinsic disorder in cellular signaling and regulation have been widely discussed , the physical basis for these proposed phenomena remains sketchy at best . An integration of multi-scale molecular modeling and experimental characterization is necessary to uncover the molecular principles that govern the structure , interaction , and regulation of IDPs . In this work , we characterize the conformational properties of two IDPs involved in transcription regulation at the atomistic level and further examine the roles of these properties in their coupled binding and folding interactions . Our simulations suggest interplay among residual structures , conformational fluctuations , and electrostatic interactions that allows efficient synergistic folding of these two IDPs . In particular , we propose that electrostatic interactions might play an important role in facilitating rapid folding and binding recognition of IDPs , by enhancing the encounter rate and promoting efficient folding upon encounter .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"physics",
"computational",
"chemistry",
"statistical",
"mechanics",
"chemistry",
"biology",
"computational",
"biology",
"biophysics",
"macromolecular",
"structure",
"analysis"
] |
2012
|
Residual Structures, Conformational Fluctuations, and Electrostatic Interactions in the Synergistic Folding of Two Intrinsically Disordered Proteins
|
The time-course of the pathological effects induced by the venom of the snake Bothrops asper in muscle tissue was investigated by a combination of histology , proteomic analysis of exudates collected in the vicinity of damaged muscle , and immunodetection of extracellular matrix proteins in exudates . Proteomic assay of exudates has become an excellent new methodological tool to detect key biomarkers of tissue alterations for a more integrative perspective of snake venom-induced pathology . The time-course analysis of the intracellular proteins showed an early presence of cytosolic and mitochondrial proteins in exudates , while cytoskeletal proteins increased later on . This underscores the rapid cytotoxic effect of venom , especially in muscle fibers , due to the action of myotoxic phospholipases A2 , followed by the action of proteinases in the cytoskeleton of damaged muscle fibers . Similarly , the early presence of basement membrane ( BM ) and other extracellular matrix ( ECM ) proteins in exudates reflects the rapid microvascular damage and hemorrhage induced by snake venom metalloproteinases . The presence of fragments of type IV collagen and perlecan one hour after envenoming suggests that hydrolysis of these mechanically/structurally-relevant BM components plays a key role in the genesis of hemorrhage . On the other hand , the increment of some ECM proteins in the exudate at later time intervals is likely a consequence of the action of endogenous matrix metalloproteinases ( MMPs ) or of de novo synthesis of ECM proteins during tissue remodeling as part of the inflammatory reaction . Our results offer relevant insights for a more integrative and systematic understanding of the time-course dynamics of muscle tissue damage induced by B . asper venom and possibly other viperid venoms .
The viperid snake Bothrops asper is responsible for most snakebite cases in Central America and some regions of Mexico and South America [1 , 2] . The local pathology induced by viperid snakes is characterized by edema , blistering , hemorrhage , lymphatic vessel damage , and necrosis of skin and muscle , some of which can be attributed to the degradation of extracellular matrix ( ECM ) [1 , 3] . Such alterations develop very rapidly after the bite , and in some cases can lead to permanent tissue damage , regardless of the application of antivenom treatment . Significant efforts have been undertaken over the last several decades to identify the toxins responsible for these effects , as well as to characterize the pathogenesis of these alterations [3–5] . Nevertheless , the complexity of this pathology demands further analyses into hitherto unknown aspects of tissue damage and the complex interplay between degenerative and early reparative events . As envenoming is a dynamic event , it is critical to investigate the process over time , which is the main focus of this study . The pathogenesis of local effects induced by B . asper venom has been studied by traditional methodologies , such as histological and ultrastructural analyses , immunohistochemical methods , and quantification of particular components and tissue markers in tissue homogenates or fluids , as a consequence of the action of crude venom and purified toxins [3 , 6–12] . Despite significant advances in the study of local tissue damage with these approaches , subtle changes in key biomarkers of tissue damage and repair during the course of envenoming remain to be identified and characterized . Moreover , since the venom is a highly complex mixture of components functioning over time , relevant information related to synergistic action of toxins could be missed when working only with isolated toxins; therefore , studies with crude venom may better advance our understanding from a predominantly reductionist to a holistic view of these multifactorial time-dependent phenomena . Recently , proteomic analysis of exudates collected around the affected tissue has become a new methodological tool to study the pathogenesis of local tissue damage induced by snake venom from a more integrative perspective [13–17] . This approach has been used to study the alterations caused by B . asper snake venom [15] , and some of its toxins , such as a myotoxic phospholipase A2 ( PLA2 ) and a hemorrhagic snake venom metalloproteinase ( SVMP ) [13 , 14 , 16] . Moreover , proteomic analysis has allowed the comparison between the action of different types of hemorrhagic and non-hemorrhagic SVMPs [13 , 16 , 17] . These studies have identified differences in the species and abundance of intracellular proteins , ECM components , and other proteins present in exudates , which offer new insights in the mechanism of action of these toxins , and in the tissue damage induced by the venom [13–16] . However , these studies have been carried only at early time periods in the course of envenoming and therefore provide only a narrow window within the whole scenario of local pathology . In the present study we analyzed the time-course variation in the protein composition and abundance of wound exudates collected from mouse gastrocnemius muscle injected with B . asper snake venom utilizing a proteomic and immunochemistry approach , in conjunction with histological analysis of tissue alterations , with the aim of identifying biomarkers of tissue damage and tissue remodeling characteristic of early and late stages of envenoming . This approach allowed the identification of key differences in some intracellular proteins and ECM components over time , which underscores the rapid cytotoxic and hemorrhagic effect of venom , followed by the action of endogenous proteinases associated with tissue remodeling later on in the course of envenoming . These results offer relevant insights for a better understanding of the complex pathological phenomena of viperid snakebite envenoming .
B . asper venom was obtained from more than 40 adult specimens collected in the Pacific region of Costa Rica and maintained at the serpentarium of Instituto Clodomiro Picado . After collection , venoms were pooled , lyophilized , and stored at -20°C until used . CD-1 mice with a body weight between 18 and 20 g were used for the in vivo studies . All the experimental protocols involving the use of animals were approved by the Institutional Committee for the Care and Use of Laboratory Animals ( CICUA ) of the University of Costa Rica ( protocol approval number CICUA 025–15 ) , and meet the International Guiding Principles for Biomedical Research Involving Animals ( CIOMS ) . Groups of four CD-1 mice ( 18–20 g ) were injected in the right gastrocnemius with 50 μg of B . asper venom , dissolved in 50 μL of 0 . 12 M NaCl , 0 . 04 M phosphate , pH 7 . 2 solution ( PBS ) . Control mice were injected with PBS alone . After 1 , 6 and 24 h of injection , mice were sacrificed by CO2 inhalation and samples of the injected muscles were resected and added to 10% formalin solution in PBS . After 48 h fixation , routine processing of tissues was performed , followed by embedding in paraffin . Sections of 5 μm thickness were obtained for each sample and stained with hematoxylin–eosin for light microscopic observation . Groups of five CD-1 mice ( 18–20 g ) were injected in the right gastrocnemius with 50 μg of B . asper venom , dissolved in 50 μL of PBS . After 1 , 6 and 24 h of injection , mice were sacrificed by CO2 inhalation , a 5 mm incision was made with a scalpel in the skin overlying the injected muscle , and a heparinized capillary tube was introduced under the skin to collect the wound exudate fluid . An approximate volume of 20–50 μL of exudate was collected from each mouse . Exudate samples were then pooled and lyophilized for further analysis . Wound exudates were collected as previously described and centrifuged at 5000 g for 3 min . The CK activity of supernatants was determined using a commercial kit ( CK-Nac , Biocon Diagnostik , Germany ) . CK activity was expressed in International Units /L ( IU/L ) . Lyophilized wound exudate samples were resuspended in water and protein quantification was performed using micro BCA protein assay kit ( Thermo Scientific ) . Twenty micrograms of protein was precipitated with acetone , resuspended in Laemmli buffer and separated in a 5–20% precast electrophoresis gel ( Bio-Rad ) . The gel was stained with Coomassie Brilliant Blue and lanes were cut into 8 equal size slices . Gel slices were destained for 3 h and the proteins were reduced ( 10 mM dithiothreitol , DTT ) and alkylated ( 50 mM iodoacetamide ) at room temperature . Gel slices were then washed with 100 mM ammonium bicarbonate , dehydrated with acetonitrile and dried in a speed vac , followed by in-gel digestion with a solution of Promega modified trypsin ( 20 ng/μL ) in 50 mM ammonium bicarbonate for 30 min on ice . Excess trypsin solution was removed and the digestion continued for 18 h at 37°C . The resulting tryptic peptides were extracted from gel slices with two 30 μL aliquots of a 50% acetonitrile/5% formic acid solution . These extracts were combined and dried to 15 μL for mass spectrometric ( MS ) analysis . LC/MS/MS was performed using a Thermo Electron Orbitrap Velos ETD mass spectrometer system . Analytical columns were fabricated in-house by packing 0 . 5 cm of irregular C18 Beads ( YMC Gel ODS-A , 12 nm , I-10-25 um ) followed by 7 . 5 cm Jupiter 10 μm C18 packing material ( Phenomenex , Torrance , CA ) into 360 x 75 μm fused silica ( Polymicro Technologies , Phoenix , AZ ) behind a bottleneck . Samples were loaded directly onto these columns for the C18 analytical runs . 7 μL of the extract was injected , and the peptides were eluted from the column at 0 . 5 μL/min using an acetonitrile/0 . 1M acetic acid gradient ( 2–90% acetonitrile over 1 h ) . The instrument was set to Full MS ( m/z 300–1600 ) resolution of 60 , 000 and programmed to acquire a cycle of one mass spectrum followed by collision-induced dissociation ( CID ) MS/MS performed in the ion trap on the twenty most abundant ions in a data-dependent mode . Dynamic exclusion was enabled with an exclusion list of 400 masses , duration of 60 seconds , and repeat count of 1 . The electrospray voltage was set to 2 . 4 kV , and the capillary temperature was 265°C . The data were analyzed by database searching using the Sequest search algorithm in Proteome Discoverer 1 . 4 . 1 against the Uniprot Mouse database from July 2014 . Spectra generated were searched using carbamidomethylation on cysteine as a fixed modification , oxidation of methionine as a variable modification , 10 ppm parent tolerance and 1 Da fragment tolerance . All hits were required to be fully tryptic . The results were exported to Scaffold ( version 4 . 3 . 2 , Proteome Software Inc . , Portland , OR ) to validate MS/MS based peptide and protein identifications , and to visualize multiple datasets in a comprehensive manner . Confidence of protein identification in Scaffold is shown as ≥ 95% confidence ( green coloration ) and 80% to 94% confidence ( yellow coloration ) . Relative quantization of proteins was performed by summing all data from the 8 gel slices for a particular sample in Scaffold and then displaying the Quantitative Value from the program . This number gives an average total of non-grouped spectral counts for a protein divided by the total non-grouping spectral counts for the 8 mass spectral runs from the gels slices from each lane ( http://www . proteomesoftware . com/ ) . The Quantitative Value allows a relative quantitative comparison between a specific protein from different samples and relative abundance between proteins for a particular exudate sample . For immunoblotting , 100 μg protein of each exudate sample were separated under reducing conditions on 4–15% Tris–HCl polyacrylamide gradient gels , and transferred to nitrocellulose membranes . Immunodetection was performed by incubating the membranes overnight at 4°C stirring with rabbit anti-type IV collagen polyclonal antibody at a dilution of 1:200 ( Abcam ab19808 ) , rabbit anti-nidogen 1 polyclonal antibody at a dilution of 1:500 ( Abcam ab14511 ) , rabbit anti-laminin polyclonal antibody at a dilution of 1:1 , 000 ( Thermo PA1-32130 ) , rabbit anti-type VI collagen polyclonal antibody at a dilution of 1:2 , 000 ( Millipore AB7821 ) , rabbit anti-type I collagen polyclonal antibody at a dilution of 1:1 , 000 ( Abcam ab21286 ) , or rabbit anti-fibronectin polyclonal antibody at a dilution of 1:3 , 000 ( Abcam ab2413 ) . The reaction was developed using anti-rabbit peroxidase antibody at a dilution of 1:5 , 000 ( Jackson ImmunoResearch ) and the chemiluminescent substrate Lumi-Light ( Roche ) . Images were captured with the ChemiDoc XRS+ System ( BioRad ) and the analysis was performed with the ImageLab software ( BioRad ) .
The pathological alterations induced by B . asper venom were studied on mouse gastrocnemius muscle tissue over a time period of 24 h . Tissue sections from control mice injected with PBS showed normal histological features of skeletal muscle tissue with transverse bundles of muscle fibers , surrounded by connective tissue and normal vascular and nerve structures ( Fig 1A ) . Tissue sections from mice injected with B . asper venom showed intense hemorrhage at 1 h ( Fig 1B ) and 6 h ( Fig 1C ) , evidenced by the presence of abundant erythrocytes in the interstitial space surrounding muscle fibers . After 24 h of injection of B . asper venom , the hemorrhage decreased since the amounts of extravascular erythrocytes was reduced as compared to previous time intervals ( Fig 1D ) , consistent with previous observations [7] . Moreover , tissue sections from mice injected with venom revealed prominent necrosis of skeletal muscle fibers at the first hour interval ( Fig 1B and 1C ) . After 24 h following venom injection , the bundles of muscle fibers appeared partially lost and disorganized with a hyaline appearance ( Fig 1D ) . These pathological observations also agree with previous studies [7 , 11 , 12] . Additionally , an infiltration of inflammatory cells was observed in tissue sections , especially after 6 h and 24 h of venom injection with a marked increment at 24 h ( Fig 1D ) . The predominant cell type was polymorphonuclear leukocytes , although a proportion of mononuclear cells , i . e . macrophages , were also observed at 24 h . These observations also agree with previous studies [9] . On the other hand , CK activity of exudate samples collected after injection of venom was 228 , 776 ± 47 , 137 IU/L at 1 h , 162 , 344 ± 23 , 371 IU/L at 6 h , and 23 , 371 ± 11 , 660 IU/L at 24 h ( Fig 1E ) . CK is a marker of plasma membrane damage and cell death of skeletal muscle fibers; hence it appears that myotoxic activity of the venom is highest at one hour , decreasing afterwards . These results are in agreement with the muscle tissue damage observed in the histological analysis , which occurs early on in the course of envenoming . Wound exudate samples collected from mice injected with B . asper venom were decomplexed by SDS-PAGE for subsequent proteomic analysis . From the mass spectrometric analysis of the gel bands , a total of 537 , 578 , and 486 proteins were identified in exudates at 1 h , 6 h , and 24 h , respectively , with protein identification probability above 95% and minimum of two peptides ( S1 Table ) . The most abundant proteins identified based on their Quantitative Value ( see http://www . proteomesoftware . com/ for full description of term ) were classified within the following groups: intracellular proteins ( Table 1 and S2 Table ) , ECM proteins ( Table 2 ) , membrane proteins ( S3 Table ) , coagulation factors ( S4 Table ) , and proteinase inhibitors ( S5 Table ) . Within each group , the proteins were organized by those that changed at least three fold as compared to another time and proteins which did not show significant change between the three time intervals , i . e . those whose amounts did not differ more than threefold between times . A total of 222 intracellular proteins ( Table 1 and S2 Table ) and 13 membrane proteins ( S3 Table ) were detected in exudates , thus demonstrating direct or indirect cellular damage induced by the venom . The most abundant intracellular proteins detected in exudates were hemoglobin subunit beta-2 and creatine kinase M-type , in agreement with the hemorrhagic and myotoxic activity of B . asper venom , respectively . Moreover , the creatine kinase M-type identified in the exudates was detected at the highest level at 1 h , and decreased over time until reaching a six fold reduction at 24 h . These results are in agreement with the CK activity of exudates and the muscle tissue damage observed in the histological analysis . In contrast , there was a trend for cytoskeletal proteins , such as actin , myosin , and tropomyosin , to increase in the exudates over time , while most of cytosolic and mitochondrial proteins appeared at the first hour of venom injection , and decreased afterwards . Of serum proteins , a total of 10 coagulation factors ( S4 Table ) and 14 proteinase inhibitors ( S5 Table ) were detected in exudates at various times . Fibrinogen beta and gamma chains appeared in the exudates at the first hour following venom injection and their abundance increased over time . Other coagulation factors detected whose amounts changed at least three fold as compared to values at other time were coagulation factor X , XII , and XIII . The inter alpha-trypsin inhibitor was the only proteinase inhibitor that increased at least threefold at 24 h as compared to 1 h and 6 h . A total of 24 ECM proteins were identified in exudates , of which 21 , 24 , and 13 proteins were detected at 1 h , 6 h , and 24 h , respectively ( Table 2 ) . Most of these proteins showed a differential abundance greater than three-fold between samples collected at different times . The most abundant basement membrane ( BM ) protein detected in the wound exudates was BM-specific heparan sulfate proteoglycan core protein ( perlecan ) , followed by alpha 1 and 2 chains of type IV collagen . Most of the BM components , such as heparan sulfate proteoglycan , type IV collagen and nidogen-2 , appeared in the exudates at 1 h , and the amount decreased over time , largely becoming undetectable at 24 h . Conversely , the amount of laminin γ-1 detected in the exudates increased at 6 h and 24 h , and the amount of nidogen-1 increased at 6 h as compared to 1 h and 24 h ( Fig 2A ) . Other collagens , such as types VI , XV , and XVIII collagens , were present in the exudates at 1 h and 6 h , but were not detected at 24 h . Interestingly , type I collagen was also detected in the exudates at the first hour and its abundance increased at 24 h ( Fig 2B ) . Other ECM proteins detected in the exudates whose abundance were greater at 6 h as compared to 1 h and 24 h were type III collagen , fibrillin 1 and 2 , chondroitin sulfate proteoglycan 4 , and type XII collagen . On the other hand , thrombospondin 1 appeared in the exudates at 1 h and decreased over time . Other ECM proteins detected in the exudates whose amounts did not vary more than threefold between times were fibronectin , thrombospondin-4 , vitronectin , dermatopontin , proteoglycan 4 , type XIV collagen , and lumican ( Table 2 ) . In order to determinate whether SVMP or endogenous proteases are active in the wound exudates , proteolytic activity assays of exudate samples were performed . Exudate samples collected from mice injected with B . asper venom at 1 h showed the highest proteolytic activity on gelatin fluorescein conjugate compared with samples collected at 6 h and 24 h . ( Fig 4A ) . When exudates collected at 1 h were incubated with polyclonal antibody against the SVMP BaP1 , the proteolytic activity of the exudates was almost completely inhibited since only 8% of the activity remained . Using zymography several gelatinolytic bands were detected corresponding to proteins of 50–150 kDa in the exudate samples collected at different times ( Fig 4B ) . An increase of two main bands of about 100 kDa and 60 kDa was observed in exudates collected at 6 h and 24 h . These molecular masses are consistent with the latent forms of matrix metalloproteinases ( MMP ) 9 and 2 , respectively [22] . Therefore , the zymography showed an increase of proteolytic activity of endogenous MMPs in exudate over time . Furthermore , bands corresponding to the molecular mass of the SVMP BaP1 were not detected in the zymographic analysis of exudate samples .
Envenoming by venomous snakes gives rise to a complex pathophysiology by virtue of the complexity of the venoms and the fact that the toxins in the venom produce manifold effects in the tissues . Proteomic analysis of wound exudates collected in the vicinity of affected tissue constitutes a powerful approach to study the pathogenesis of tissue damage induced by snake venoms from a more comprehensive perspective [13–17] , thus complementing histological , ultrastructural and biochemical analyses . This methodological tool has been used to investigate the early alterations provoked by B . asper venom [15] and some of its toxins , especially myotoxic PLA2s and hemorrhagic SVMPs [13 , 14 , 16] , as well as the inhibitory effects of antivenoms and low molecular mass inhibitors [15] . However , analyses in these previous studies were performed at a single time interval after injection , thus precluding the understanding of these events from a time-course perspective . In this study , we investigated the dynamics of local effects induced by B . asper venom in the gastrocnemius muscle of mice at various time intervals . Our histological and biochemical observations agree with previous studies showing a rapid development of myonecrosis and hemorrhage , followed by an inflammatory process characterized by the infiltration of neutrophils and macrophages at later time intervals [4 , 5 , 7 , 9 , 10] . Previous works on venom-induced myonecrosis have quantified CK activity in plasma [6] , where the highest levels were observed at 3 h post envenoming . In contrast , in exudates , highest CK levels are higher than in plasma , and peak activity occurs at 1 h instead of 3 h . This difference may be attributed to the kinetics of absorption of this enzyme into the circulation after its release from damaged muscle fibers , since exudate was collected close to the venom-injected muscle . In agreement with previous pathological and proteomic studies , intracellular proteins were abundant in exudates , as a consequence of the cytotoxic effect of venom , especially on skeletal muscle fibers [7 , 11 , 14 , 15] . The time-course analysis of the intracellular proteins in exudate underscores that most of the cytosolic and mitochondrial proteins appear early on due to the rapid action of myotoxic PLA2s and PLA2 homologues in muscle tissue [7 , 11 , 14 , 15 , 23] , followed by a decrease of these proteins . Most of these proteins are derived from the cytosol of skeletal muscle fibers since myotoxic PLA2s induce a rapid disruption of the integrity of muscle cell plasma membrane [4 , 7 , 24] . The high CK activity of exudates at 1 h and our histological observations corroborate the early onset of myonecrosis in the course of envenoming and agree with proteomic analysis . In contrast to cytosolic proteins , most of the cytoskeletal proteins , such as actin , myosin , and tropomyosin , are more abundant in exudates collected at later time periods . This late increment suggests that the presence of cytoskeletal protein fragments in the exudate depends on the action of proteinases that release these structural components from damaged cells . A prominent calcium influx in muscle cells occurs after venom-induced plasma membrane damage [25 , 26] . An increased calcium concentration in the cytosol results in the activation of calpains , which might hydrolyze cytoskeletal components [27] . Subsequently , proteinases derived from inflammatory cells arriving at the necrotic tissue may also contribute to proteolysis of muscle cytoskeletal proteins [9 , 10] . Thus , the proteomic analyses reveals two ‘waves’ of release of intracellular proteins to exudates: an early release of cytosolic and mitochondrial proteins , which depends on the rapid myotoxin-induced membrane damage , and a more delayed release of cytoskeletal protein fragments , which is due to proteolytic degradation . The presence of cell membrane-associated proteins may be evidence of direct or indirect cellular damage induced by the venom . Moreover , proteolysis of these components , either by venom or endogenous proteinases , may cause their ‘shedding’ and diffusion to the exudate compartment . It is tempting to speculate that , in addition to being a passive reflection of venom-induced plasma membrane damage , the release of these protein fragments may also play a functional role in cellular signaling associated with inflammatory and reparative events . However , the pathological relevance of the hydrolysis of these proteins in the overall mechanism of local tissue damage induced by snake venoms has not been established and needs further study . The presence of ECM proteins in wound exudates reflects the cleavage by either venom-derived proteinases or endogenous proteinases , such as MMPs , generated in the course of the inflammatory response . The degradation of ECM is a relevant component of viperid venom-induced tissue damage , and proteomic analysis has been particularly useful in revealing a complex pattern of hydrolysis [14–16] . Previous studies detected B . asper venom components in muscle homogenates of mice during the first week after experimental envenoming [28]; however , the activity of these toxins has not been previously addressed . When assessing the proteinase activity of exudates , highest activity was detected in samples collected after 1 h of envenoming; here we demonstrate that this is mainly due to the action of SVMPs , since antibodies against BaP1 , the most abundant proteinase in B . asper venom [29] , almost fully inhibited exudate-induced proteolysis . However , this enzymatic activity decreased over time , probably as a consequence of diffusion of venom components from the injected muscle or of inhibition by plasma or tissue-derived proteinase inhibitors . It is likely that activity at later time intervals , i . e . 24 h , is mostly due to endogenous MMPs generated in the course of the inflammatory response , such as MMP-9 and MMP-2 , which was confirmed by the detection with zymography of the wound exudates , although it remains possible that some venom proteinases persisting in the tissue may also contribute to this observation . These results agree with previous studies which demonstrated an increase in the expression of MMP-9 and MMP-2 in muscle tissue injected with B . asper venom [28] or with purified SVMP and PLA2 toxins [30] . Taken together these findings suggest that the hydrolysis of ECM is mainly due to SVMPs in the early stages of envenoming , while endogenous MMPs participate later in the course of envenoming . A large body of experimental evidence indicates that BM and related ECM components that provide stability to microvessel structure are the key targets of hemorrhagic SVMP [5 , 13 , 14 , 16 , 31–34] . Moreover , SVMP-induced hemorrhage occurs very fast after injection [12 , 35–37] . Therefore , the presence of ECM components in wound exudates during first hour as compared to later time periods may offer important insights for understanding the mechanism of action of hemorrhagic SVMPs . Regarding BM components , the presence of degradation products of perlecan , type IV collagen , nidogen , and laminin in wound exudates underscores a rapid and drastic damage of BM structure . In particular , perlecan and type IV collagen are abundant in exudates after the first hour , when hemorrhagic events have occurred , and then their amounts decrease over time , as they were not detected by proteomic analysis at 24 h . Western blot analysis of exudate confirmed the presence of fragments of type IV collagen 1 h after venom injection . Perlecan is the most abundant BM protein detected in the wound exudates during the first hour . In previous proteomic studies , we have found that the relative amount of perlecan in wound exudates induced by a hemorrhagic SVMP was greater as compared to a non-hemorrhagic one [13] , but similar when compared to other hemorrhagic SVMPs [16] , and its presence was abolished when B . asper venom was previously incubated with batimastat [15] , a metalloproteinase inhibitor . Such findings , together with our data , suggest that degradation of perlecan in early stages of envenoming may play an important role in the hemorrhagic mechanism of SVMPs . This proposal agrees with the known structural role of perlecan in BM [38–42] . In addition , mutations in the perlecan gene in mice have been associated with loss of BM integrity in different tissues [43–45] , including microvasculature of brain and skin , which cause severe bleedings due to dilatation and rupture of microvessels [45] . On the other hand , previous proteomic studies using similar models have not detected type IV collagen in wound exudates induced by B . asper venom or its toxins [13–16] . However , several studies using in vitro and in vivo models have identified type IV collagen as one of the most likely key components degraded by hemorrhagic SVMPs and associated with the initial microvessel destabilization and hemorrhage [5 , 13 , 16 , 34 , 46 , 47] . Our present proteomic results did identify fragments of type IV collagen in exudates at 1 h , thus agreeing with previous immunohistochemical and immunochemical evidence [13 , 16] . In addition , according to Western blot analysis , degradation products of type IV collagen appear in exudates in samples collected at 1 h . Such early appearance of type IV collagen and perlecan strongly suggest that their degradation is due to the direct proteolytic activity of SVMPs . The hypothesis that type IV collagen is a key target in the hemorrhagic mechanism of SVMPs is compatible with the structural role of this collagen in the mechanical stability of BM , as it is stabilized by covalent cross-links [41 , 42 , 48–51] . In addition , mutations on type IV collagen genes have been associated with pathological alterations in microvessels and with hemorrhage in brain , kidney and lungs in mice and humans [52–58] . Thus , the rapid hydrolysis of perlecan and type IV collagen after injection of B . asper venom supports the view that BM destabilization leading to hemorrhage is likely to depend on the degradation of these mechanically-relevant components . Nidogen 2 appeared in early time periods in wound exudates , in agreement with previous proteomic studies [13 , 15] , and then it decreased over time in our proteomics analysis . Since nidogen 2 is more abundant in the BM of blood vessels [59] , and its time-course dynamics of appearance in exudates is similar to that observed for type IV collagen and perlecan , the release of nidogen 2 might be associated with vascular BM damage . In contrast , taken together the proteomic and Western blot analyses showed that nidogen 1 increased over time in wound exudates . In addition , nidogen 1 and 2 have been detected in plasma of healthy mice [60] , which could explain the presence of nidogen 1 in the wound exudates according to Western blot results . On the other hand , laminin γ1 , which is widely distributed [61 , 62] , also increased over time in wound exudates . The time-course variation of the molecular masses of immunoreactive bands in the cases of nidogen 1 and laminin underscores the dynamics of degradation of these components over time . Furthermore , Escalante et al . [13] demonstrated similar patterns of degradation for nidogen and laminin in muscle tissue induced by hemorrhagic and non-hemorrhagic SVMPs . Our observations allowed the analysis of the time-course dynamics of the hydrolysis of non-fibrillar collagens associated with the BM , such as types VI , XV , and XVIII collagens . As in the case of type IV collagen and perlecan , hydrolysis of these components was highest at 1 h , hence indicating a rapid degradation , probably by venom proteinases . These collagens connect the BM with fibrillar collagens of the matrix [39 , 63] , and are known to play a relevant role in the mechanical stability and integration of BM with connective tissue [39 , 63] . Hence , the hydrolysis of these components by SVMP might be also critical for capillary wall destabilization , as have been previously proposed [13 , 64] . Alternatively , the increase of these collagens in exudates might be consequence of BM damage after the hydrolysis of other components , such as type IV collagen and perlecan . Type VI collagen is more abundant in the BM of muscle cells [65–67]; thus the increment of non-degraded type VI collagen chains in exudates could reflect synthesis de novo during reparative and regenerative events in muscle tissue . The role of the degradation of these collagens in the initial destabilization of BM induced by hemorrhagic SVMP is an issue that should be further investigated . Other ECM components of interest detected in the proteomic analysis are collagen I and fibronectin . Collagen I is a fibril-forming collagen distributed in non-cartilaginous connective tissues such as skin and connective tissue of muscle [63] . According to proteomic results , the relative abundance of collagen I in exudates is higher at 24 h as compared to 1 and 6 h . This late hydrolysis of collagen I could be result of the action of endogenous MMPs synthesized during the course of inflammation in the damage tissue . Fibronectin was detected in the exudates both in proteomic and immunochemical analyses . This protein can be found in two forms: plasma fibronectin , which is a soluble molecule synthesized by hepatocytes , and cellular fibronectin , which is produced in the tissues and is incorporated in the ECM [62] . Thus , the presence of fibronectin in exudates could be either a consequence of plasma exudation or hydrolysis from the ECM . According to proteomic analysis , the amount of fibronectin in exudates does not change over time; however , on the basis of Western blot analysis , it appears to be more degraded at early time periods most likely due to the action of SVMPs . Taken together , our observations highlight a dual pattern of ECM protein degradation and appearance in exudates . Types IV and VI collagens , perlecan , nidogen and fibronectin show a higher degradation early on in the course of envenoming , correlating with the rapid action of SVMPs upon venom injection , as demonstrated by the proteinase activity of exudates . The rapid action of SVMPs on various key components of the BM is likely to be causally related to microvessel damage and hemorrhage . In the case of the fibrillar collagen I , it seems to be degraded predominantly by endogenous MMPs at later time periods , during the inflammatory reaction that ensues in the tissue as a consequence of venom-induced damage , as evidenced by zymography . The presence of abundant plasma proteins in the exudate , as revealed by proteomic analysis , is a consequence of plasma exudation as a result of edema and increment in vascular permeability induced by the venom [68 , 69] . Some of the plasma proteins detected are acute-phase proteins , proteinase inhibitors and coagulation factors . Of interest is the increase of fibrinogen and the inter α-trypsin inhibitor heavy chains over time . The presence of fibrinogen in exudates might be secondary to the inflammatory exudation induced by the venom since this protein is typically found in plasma at high concentrations [70] . Previous proteomic studies have found fibrinogen in the wound exudates induced by B . asper venom and its toxins , especially SVMP BaP1 , early in the course of envenoming [14 , 15] . Our data show an increase of fibrinogen in wound exudates over time . This increment might be consequence of fibrin clot formation in capillary walls , due to vascular damage induced by SVMPs [12 , 14 , 35] , and also to fibrin formation in the extravascular interstitial space , followed by fibrinolysis by endogenous proteinases [70] , thus explaining their higher amounts in exudates collected at later time intervals . The inter-α-trypsin inhibitor heavy chains are mainly secreted into the blood by the liver as serum protease inhibitor whose concentration increases in inflammatory conditions [71] . The effect of these proteins in tissues has been associated with both inflammatory and anti-inflammatory activities [71–73] . Moreover , these proteins can be covalently linked to hyaluronan , exerting functions on cell migration and ECM remodeling under physiological and pathological conditions [74 , 75] . Thus , the increase of inter-α-trypsin inhibitor heavy chains detected in exudates might be due to an acute-phase inflammatory response and to the tissue inflammation as a consequence of venom-induced damage . In conclusion , the proteomic analysis of wound exudates performed in this study provides a more complete understanding of the time-course dynamics of muscle tissue damage induced by B . asper venom . These observations , together with Western blot and histology data , provide a more integrated view of venom-induced local tissue damage ( Fig 5 ) . The early presence of cytosolic and mitochondrial proteins in exudates , as compared to the later increase of cytoskeletal proteins , confirms the rapid cytotoxic effect of venom , followed by the action of endogenous proteinases in the cytoskeleton of damaged muscle fibers . On the other hand , the early presence of BM and other BM-associated ECM components in exudates , together with venom-derived proteolytic activity of exudates , strongly suggest the hydrolysis of these components by SVMPs in the early stages of envenoming . In contrast , the increment of some ECM proteins in the exudates at later time intervals is likely to be due to the action of endogenous MMPs or to their synthesis de novo during tissue remodeling associated with inflammation and reparative processes . Finally , the time-course of appearance in wound exudates of type IV collagen and perlecan supports the role of the hydrolysis of these BM components in the mechanism of microvascular damage induced by hemorrhagic SVMP .
|
The local pathology induced by viperid snakes is characterized by a complex of alterations as consequence of direct and indirect effects of the toxins present in the venom , as well as the host response to tissue damage , and constitutes a dynamic process of degenerative and reparative events . The pathogenesis of local effects induced by Bothrops asper venom has been studied by traditional methodologies . Recently , proteomic analysis of wound exudates collected in the vicinity of affected tissue has become a powerful tool to study the pathogenesis of local envenoming from a more integrative perspective . Thus , in the present study we analyzed the dynamics of the local effects induced by B . asper venom in the gastrocnemius muscle of mice through a proteomic and immunochemistry approach in order to identify biomarkers of tissue damage and repair during the course of envenoming . Our results showed an early presence of cytosolic and mitochondrial proteins in exudates as compared to cytoskeletal proteins , which reflect the rapid cytotoxic effect of venom , followed by the action of endogenous proteinases in the cytoskeleton of damaged muscle fibers later on in the course of envenoming . On the other hand , the early presence of extracellular matrix components and the increment of some of them in exudates , reflect the rapid microvascular damage and hemorrhage induced by the venom , followed by the action of endogenous matrix metalloproteinases ( MMPs ) during tissue remodeling as part of the inflammatory response . Overall our study allowed the identification of key biomarkers of tissue damage and repair as part of the pathological effects induced by B . asper venom in skeletal muscle , which offer relevant insights for a better understanding of the complex dynamics of local pathology induced by viperid snakebite envenoming .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"toxins",
"classical",
"mechanics",
"pathology",
"and",
"laboratory",
"medicine",
"enzymes",
"enzymology",
"toxicology",
"collagens",
"toxic",
"agents",
"membrane",
"proteins",
"damage",
"mechanics",
"cellular",
"structures",
"and",
"organelles",
"venoms",
"proteins",
"proteomics",
"cell",
"membranes",
"physics",
"biochemistry",
"cytoskeletal",
"proteins",
"cell",
"biology",
"biology",
"and",
"life",
"sciences",
"proteases",
"physical",
"sciences",
"extracellular",
"matrix",
"proteins"
] |
2016
|
Muscle Tissue Damage Induced by the Venom of Bothrops asper: Identification of Early and Late Pathological Events through Proteomic Analysis
|
The recent spread of Zika virus ( ZIKV ) and its association with increased rates of Guillain Barre and other neurological disorders as well as congenital defects that include microcephaly has created an urgent need to develop animal models to examine the pathogenesis of the disease and explore the efficacy of potential therapeutics and vaccines . Recently developed infection models for ZIKV utilize mice defective in interferon responses . In this study we establish and characterize a new model of peripheral ZIKV infection using immunocompetent neonatal C57BL/6 mice and compare its clinical progression , virus distribution , immune response , and neuropathology with that of C57BL/6-IFNAR KO mice . We show that while ZIKV infected IFNAR KO mice develop bilateral hind limb paralysis and die 5–6 days post-infection ( dpi ) , immunocompetent B6 WT mice develop signs of neurological disease including unsteady gait , kinetic tremors , severe ataxia and seizures by 13 dpi that subside gradually over 2 weeks . Immunohistochemistry show viral antigen predominantly in cerebellum at the peak of the disease in both models . However , whereas IFNAR KO mice showed infiltration by neutrophils and macrophages and higher expression of IL-1 , IL-6 and Cox2 , B6 WT mice show a cellular infiltration in the CNS composed predominantly of T cells , particularly CD8+ T cells , and increased mRNA expression levels of IFNg , GzmB and Prf1 at peak of disease . Lastly , the CNS of B6 WT mice shows evidence of neurodegeneration predominantly in the cerebellum that are less prominent in mice lacking the IFN response possibly due to the difference in cellular infiltrates and rapid progression of the disease in that model . The development of the B6 WT model of ZIKV infection will provide insight into the immunopathology of the virus and facilitate assessments of possible therapeutics and vaccines .
Zika virus ( ZIKV ) is an emerging mosquito-borne pathogen that belongs to the Flavivirus genus of the Flaviviridae family , which includes globally relevant arthropod-transmitted human pathogens such as dengue ( DENV ) , yellow fever ( YFV ) , West Nile ( WNV ) , Japanese encephalitis ( JEV ) , and tick-borne encephalitis viruses . The first strain of ZIKV ( MR 766 ) was isolated in 1947 from a febrile sentinel rhesus monkey in the Zika forest near Entebbe , Uganda after the virus underwent intracerebral passage in Swiss albino mice[1] . For the next 50 years infections with ZIKV were reported sporadically in different regions of Africa and Asia , but were associated with mild symptoms consisting of skin rashes , conjunctivitis , fever and headaches[2] . In 2007 ZIKV started spreading west , first with an outbreak in Island of Yap where it infected over 70% of the population , followed in 2013 by an outbreak in French Polynesia[3] . This last outbreak was associated with a sharp increase in cases of Guillain Barre Syndrome ( GBS ) , an autoimmune disease characterized by weakening and even paralysis of the limbs and face [4 , 5] . In 2015 Zika spread to South and Central America , infecting thousands of people in Brazil and Colombia , where it associated with an increase in GBS rates as well as a significant increase in severe fetal abnormalities that include spontaneous abortion , stillbirth , hydrocephaly , microcephaly , and placental insufficiency[6–9] . The temporal association of the viral outbreak and increased incidence of GBS and birth defects did not necessarily imply causality , however recent studies showing infection of the CNS in utero as well as infections of human neural progenitor and neural stem cells leading to cell cycle arrest and death [10 , 11] lend credence to the causality of ZIKV in microcephaly [12 , 13] . While a majority of ZIKV infections in adults are asymptomatic or result in skin rashes , conjunctivitis ( nonpurulent ) , muscle pain and joint pain ( small joints of hands and feet ) , or a headache lasting for 2–7 days , the recent spread of ZIKV and its association with increased rates of neurological disorders has created an urgent need for animal models to examine the pathogenesis of the disease and explore the efficacy of potential therapeutics and vaccines . In recent weeks , studies showed that young or adult immunocompetent mice are not susceptible to infection , but prepubescent mice lacking the capacity to produce or respond to interferons ( IFNs ) , including A129 ( type I IFNAR KO ) , Interferon Regulatory Factor ( IRF ) 3/5/7 triple KO , and AG129 ( type 1 and type 2 IFN KO ) , develop neurological disease and succumb to infection with high viral loads in the brain , spinal cord , and testes [14 , 15] . Interestingly , in seeking a model that was less dramatically challenged Lazear et al infected mice deficient in IRF3 , IRF5 , or Mitochondrial Anti-Viral Signaling ( MAVS ) , which mediate the signaling of pattern recognition receptors for ssRNA RIG-I and MDA5 , however none of these mouse models developed disease [14] . While these IFN deficient models provide useful data , the profound immunological defects in these strains may skew our understanding of the pathophysiology of the disease as the impaired IFN response can , for example , modify the susceptibility to infection of specific tissues . Moreover , previous studies on the pathology of flavivirus suggest that pathogenicity may be determined not just by the effect of the virus but by the immune response it elicits [16] . Thus an immunocompetent mouse model is urgently needed to understand the host response and pathogenesis of the disease and to test and compare the potency of potential therapeutic approaches . More recently several studies have shown that mice from immunocompetent strains can be infected in utero provided a high titer infection is achieved in the dam or antibodies neutralizing interferon are administered [17–20] . These infections , when performed early in pregnancy ( E5-E8 ) , result in increased fetal resorption and altered brain and eye development . The relative contribution of virus-induced placental insufficiency versus direct deleterious effect of the virus on the cells of the fetal CNS is unclear [20 , 21] . In mice , the stage of CNS development of neonatal pups has been equated to a human mid-term fetus [22] . Neonatal rodents have shown to be highly susceptible to many neurotropic viral infections , including Herpes , Bornavirus , Tacaribe arenavirus and more recently Chikungunya , that present with meningoencephalitis [23–26] . In this study we establish a new model of subcutaneous ( SC ) ZIKV infection in neonatal ( 1 day old ) , immunocompetent C57BL/6 ( B6 WT ) mice and compare its clinical progression , virus distribution , immune response , and neuropathology with C57BL/6-IFNAR-/- ( IFNAR KO ) mice , which are deficient in type 1 IFN responses . We show that immunocompetent mice , when infected at day 1 of age ( P1 ) develop unsteady gait , loss of balance , kinetic tremors , severe ataxia and seizures beginning around 13 days post infection ( dpi ) that subside 2 weeks later . Infection-induced IFN responses appear to reduce but not completely abrogate CNS infection in B6 WT mice . Further , whereas the response to the virus in the CNS of B6 WT mice is characterized by cellular infiltration consisting predominantly of CD8+ T cells and is associated with increased expression levels of T cell effector molecules such as IFNg , granzyme B and perforin , the CNS in IFNAR KO mice show infiltration predominantly by neutrophils and macrophages as well as higher levels of inflammatory cytokines . Lastly the CNS of B6 WT mice shows evidence of neurodegeneration that is less prominent in IFNAR KO mice . This model does not address transplacental transmission of virus from mother to fetus or adult transmission , however it offers an immunocompetent symptomatic mouse model for ZIKV infections that may prove useful to understand the long term effects of ZIKV infection . In addition , it avoids transplacental infection and consequent placental insufficiency as a confounding factor in the development of the brain . Lastly , this model may help understand the clinical consequences for infections in late pregnancy or early childhood . While ZIKV infections in early pregnancy have catastrophic consequences , there is concern that infections at later stages of gestation or early childhood may result in long term neurodevelopmental issues that we are as yet unaware of .
Recent studies showed that mice defective in interferon responses are susceptible to infection and develop a lethal disease . To explore whether IFNAR KO mice are susceptible to infection with the contemporary ZIKV PRVABC59 strain , 10 day old ( P10 ) IFNAR KO mice were challenged subcutaneously ( sc ) with 2 x 103 PFU of ZIKV . The mice remained asymptomatic and maintained their weight gain for the first 4 dpi ( Fig 1B ) . Beginning late on 4 dpi the mice demonstrated reduced movement , tremors , bilateral hind limb paralysis , and died within 24 hours of disease onset ( Fig 1 ) . These results were consistent with those reported for 3 week old: A129 , IRF-3/5/7 -/- and AG129 ( IFNa/b/g KO ) mice challenged with the African MR766 or MP1751 strains , or the more recent H/PF/2013 strain [27–29] . We next examined whether the ZIKV PRVABC59 strain could be used to challenge B6 WT mice . Reports on susceptibility to infection in utero[30 , 31] , ZIKV’s ability to infect developing neurons in vitro [10 , 17] and prior studies in TCRV[25 , 32] , Sindbis ( manuscript in preparation ) , and Chikungunya [26] suggested that a challenge with ZIKV very early in life , when the central nervous and immune systems are not fully mature may yield a productive infection . Thus we explored whether B6 WT mice were susceptible to ZIKV if challenged one day after birth ( P1 ) . As shown in Fig 1 , WT mice infected on P1 with PRVABC59 ( 2x103 PFU ( sc ) ) remain asymptomatic for 12 days , except for a decreased rate of weight gain . After 2 weeks the mice develop unsteady gait with widening stance , hyperactivity and ataxia . This is followed by reduced mobility , intermittent alternating collapse of the hind limbs , loss of balance and seizures ( S1–S3 Movies ) . Interestingly , unlike IFNAR KO mice , these mice do not develop flaccid hind limb paralysis or succumb to the infection ( Fig 1 and S4 Movie ) . Indeed , the observed symptoms diminish over the course of 2 weeks and most mice survive the challenge and recover . Additional studies will be needed to assess the long term consequences of infection . To determine whether the difference in clinical presentation was due to the age of the mice at the time of challenge , IFNAR KO mice were challenged at P1 and P3 . The clinical presentation in mice challenged on P1 or P3 was similar to that of mice challenged on P10 as they developed bilateral paralysis and succumbed to disease by 5 dpi ( S1 Fig ) . Conversely , B6 wt mice challenged on P3 or P10 do not develop signs of disease . This indicated that the virus was inducing fundamentally different pathology in immunocompromised and immunocompetent mice . Given that the clinical development of the disease was similar for mice challenged on P1 , P3 or P10 , all the ensuing experiments used P10 infections for the IFNAR KO model so that the peak of disease coincides in age with that of the B6 model ( P15 ) . Previous reports show that in mice defective in IFN responses ( A129 , AG129 , IRF-3/5/7 triple KO ) the virus distributes systemically , with detectable ZIKV present in the brain and spinal cord , testes , spleen , liver , kidney and serum [28] . Our challenge model with PRVABC59 in IFNAR KO mice confirms these results showing high virus titers in the CNS ( 4 . 4 x 107 TCID50 ) , as well as spleen ( 9 x 105 TCID50/0 . 5 g of tissue ) , and liver ( 1 . 5 x 105 TCID50/0 . 5 g of tissue ) at 5 dpi ( Fig 2A ) . In comparison , WT mice challenged at P1 show relatively lower viral loads in the CNS ( 9x104 TCID50/0 . 5 g of tissue ) at 15 dpi , the time when the animals displayed peak neurological deficit . Moreover , the B6 WT mice do not show evidence of viral infection in spleen or liver , indicating selective infection of the CNS . Similarly , quantitative real-time PCR showed detectable levels of viral RNA in the CNS of B6 WT mice starting at day 3 and increasing through day 9 of infection ( Fig 2B and 2C ) . Of note the levels of viral RNA in CNS did not reach those evident in IFNAR KO mice at peak of disease ( 105 ZIKV RNA copies/mL vs 108 ZIKV copies/mL ) ( Fig 2A and 2B ) . No RNA was detectable in liver or spleen of B6 mice ( Fig 2B ) . The presence of more than 105 ZIKV RNA copies /mL in liver and spleen of IFNAR KO mice suggests that type I IFNs play a key role in controlling the virus in the peripheral organs and peripheral infection may play a role in lethality in these models . In contrast , infection in B6 WT appears to be restricted to the CNS possibly due to lower levels of IFNs in response to the virus in CNS as previously shown [33 , 34] . Given that both strains were positive for ZIKV in the CNS but displayed different clinical presentation , we next explored whether the distribution of the virus in the CNS was similar in B6 WT and IFNAR KO mice . Immunohistochemistry of the CNS using a mouse monoclonal antibody ( clone D1-4G2-4-15 ) that has been shown to react with ZIKV and other members of the flavivirus family , demonstrated detectable viral antigen in the CNS of both B6 WT ( 15 dpi ) and IFNAR KO ( 5 dpi ) mice ( Fig 2D ) . As expected , age-matched , uninfected control sections were negative for the virus stain ( Fig 2D ) . Consistent with the higher virus titers , IFNAR KO mice showed stronger and broader staining of the virus by immunohistochemistry that extended to the cortex . In contrast , the CNS of B6WT mice showed virus predominantly in the cerebellar white matter and granular layers as well as in the hippocampus region ( S2 Fig ) , but not in the frontal cortex . Colocalization of stains for ZIKV and neurofilament heavy chain in the areas of cerebellum ( Fig 2E ) and hippocampus ( S2 Fig ) indicates that ZIKV infects neurons . We next determined whether the immune response to the virus in the CNS was similar between IFNAR KO and B6 WT mice . Both strains showed profound up-regulation of the expression of genes linked to inflammation and cellular infiltration . These included a significant up-regulation of Ccl2 , Ccl5 , Cxcl10 and Cxcl11 with the corresponding increases in genes linked to the recruitment and activation of neutrophils ( PMN ) and monocyte/macrophages including increases in MHC , Cd80 , Cd86 , Cd68 and Cd40 as well as marked increases in Csf1 and Csf2 ( Fig 3 ) . These were accompanied by significant increases in the expression of IFNb , Tnfa , Il6 , Il1 , Ifng , C3 and Cox2 , all indicating a severe inflammatory response in the CNS . In all , 49 of the 96 genes screened showed at least a 5 fold increase in expression in both strains . Interestingly , the magnitude of the up-regulation for several of these genes was markedly different between the models . For example , IFNAR KO mice showed significantly higher levels of Csf2 , Csf3 , Sele and Selp , while B6 WT mice showed relatively higher levels of genes linked to antigen presentation such as H2-Eb1 and B2m . B6 WT mice also showed significantly higher levels of Cd45 , Ccr7 and Cxcr3 , suggesting increased infiltration of peripheral leukocytes , while IFNAR KO mice showed higher levels of Ccr4 likely expressed on microglia and astrocytes ( Fig 3 ) . Among the genes linked to inflammation , the expression of Ifna , Ifnb , Cox2 , Il1 , and Il6 were significantly higher in the IFNAR KO mice potentially due to higher virus titers in the CNS and the deficient IFN response . In contrast , the CNS of B6 WT mice had relatively higher expression of ISGs OAS1 and ISG15 , as well as genes corresponding to T cells including CD3 , CD4 and CD8 , and markers of Th1 and cytolytic responses such as GzmB and Prf1 , Il2 , Ifng , and STAT1 . These data would indicate that ZIKV enables a significant T cell response in the CNS of B6 WT mice that is not evident in IFNAR KO mice . The data above suggested that the CNS infection in both strains was accompanied by microglial/macrophage activation and immune cell infiltration . To explore this further , we isolated cells from the CNS of infected animals at 15 dpi and studied the cell populations using flow cytometry . As predicted by the increased expression of genes related to T cells and antigen presenting cells , there was significant cellular infiltration of CD45hi cells in the CNS of both strains . In B6 WT mice the majority of infiltrating cells were T cells , with CD8+ T cells comprising 45% and CD4+ T cells comprising 20% of the CD45hi infiltrating population . The remaining cell types consisted of F4/80+CD11b+ macrophages ( 15% ) , NK1 . 1+ Natural Killer cells ( 3% ) and CD19+CD45R+ B cells ( 5% ) ( Fig 4A ) . IFNAR KO mice showed even higher levels of cellular infiltration , however in these mice the infiltrating cells corresponded to CD11b+Ly6G+ and CD11b+F4/80+ consistent with PMN and macrophages respectively , with only a minor population of T cells and NK cells ( Fig 4A ) . These data together with the analysis of gene expression suggests that the inflammatory and immune processes that follow ZIKV are fundamentally different in IFNAR KO and B6 WT mice , with IFNAR KO mice showing significant inflammation of the CNS accompanied by infiltration by PMN and granulocytes , while IFN-sufficient animals mount a T cell driven response characterized by CD8+ T cells and high levels of IFNγ and granzyme B . Previous studies had shown inflammatory and degenerative changes in the brains of IFN deficient mice challenged with ZIKV . These include the presence of scattered nuclear fragments , perivascular cuffing , and PMN infiltrating gray and white matter [27] . To determine whether the B6 WT mice would show similar evidence of inflammatory and degenerative changes , we stained sagittal brain sections collected from Zika-infected B6 WT mice at 15 dpi and age-matched control animals . Fluorescence immunohistochemistry confirmed the presence of CD45+ immune cells in the parenchyma of the CNS in both B6 WT and IFNAR mice and showed that these cells concentrate in the white matter and granular layers of the cerebellum ( Fig 4B ) , consistent with previous studies in IFN deficient mice [27] . The clinical presentation of ZIKV infected B6 WT mice , along with the previously described tropism of ZIKV for neurons suggested that the virus infects and damages the CNS of B6 WT mice . To test this , we stained sections adjacent to those used to detect virus and inflammation with the Fluoro-Jade C , a stain that specifically labels degenerating neurons . As expected , uninfected , age-matched controls showed no staining with the Fluoro-Jade C stain . In contrast , infected B6 WT mice showed foci of Fluoro-Jade C positive neurons in all layers of the cerebellum , but predominantly in the granular and PC layers ( Fig 5 ) . Interestingly , there were fewer Fluoro-Jade C+ cells in granular and Purkinje layers of the cerebellum of IFNAR KO mice . The presence of infiltrating CD8+ T cells and higher levels of fluorojade C positive cells in the CNS of B6 WT mice suggests a possible role for CD8+ T cells in the pathology .
ZIKV belongs to the Flavivirus genus that includes several etiological agents of viral encephalitis , the most significant being Japanese encephalitis virus , West Nile virus , and tick-borne encephalitis virus . As with other flaviviruses , the majority of infected individuals will not develop disease , but a minority will develop a severe illness with a significant chance of permanent neurological damage , congenital malformations , or death . The factors that determine this are likely numerous , involving complex interactions between virus and host that are yet to be uncovered . Animal models can help us understand the pathophysiology of the virus , identify therapeutic targets , and explore the safety and efficacy of new therapeutics and vaccines . This study shows that neonatal B6 WT mice challenged with ZIKV develop a slow onset non-lethal encephalitis that is characterized by unsteady gait , kinetic tremors , severe ataxia , loss of balance and seizures . The virus localizes to the CNS where it elicits a strong IFN response , T cell infiltration with increased expression of RNA coding for Ifng , granzymeB , perforin1 and Il-2 . In addition , these mice show evidence of neurodegeneration in particular affecting the Purkinje and granular cell layers of the cerebellum as evidenced by Fluoro-Jade C staining . Our data suggests that innate and adaptive responses can limit viral expansion but may also play a role in pathological changes in the CNS . In response to the outbreak of ZIKV and its association with increased congenital and neurological disease , animal models are being developed with unprecedented celerity to understand the pathogenesis of the disease and test possible therapeutics and vaccines . Early studies in mice had suggested that ZIKV can replicate and cause injury in cells of the central nervous system [1 , 35 , 36] but used the prototype MR 766 strain of ZIKV , which had undergone extensive passage in suckling mouse brains . Several new mouse models of ZIKV were developed over the past 6 months using current viral isolates . Using a low passage Cambodian isolate of ZIKV , Rossi et al showed that 3 week old A129 and AG129 mice , which lack type I or I & II IFN respectively , develop paralysis and succumb to disease by 7 dpi while older mice showed viremia and weight loss but recovered after day 8[15 , 29] . Similar results were observed by Lazear et al , using either 4–6 week old A129 or Irf3/5/7 triple knockout mice challenged with ZIKV ( H/PF/2013 ) from French Polynesia , as well as African ZIKV strain MR 766[28] . Our studies advance on these findings by characterizing the immune response to ZIKV in the CNS and key peripheral organs . We show that in the CNS of mice with deficient IFN responses the virus in the brain was localized predominantly to the cerebellum at the peak of disease and elicits a marked inflammatory response characterized by significant increases in the mRNA expression of complement ( C3 ) , Cox2 , Il1a , Il1b , and Il6 . The increased mRNA expression of Sele , Selp , Csf2 and Csf3 is consistent with the observed increase in infiltrating neutrophils and macrophages evident by flow cytometry . Together this pattern of expression suggests the activation of microglia and/or infiltration of activated macrophages , which has been shown to lead to uncontrolled inflammation and neuronal death in other models of flavivirus encephalitis such as mice infected with Japanese encephalitis virus [37] . Interestingly , despite significant upregulation of markers for inflammation and the evidence of infiltrating neutrophils and macrophages , the infected IFNAR KO mice did not show significant increase in the expression of genes linked to apoptosis ( including BCL2 , Bax , Agtr2 or Bcl2l1 ) ( S3 Fig ) . This differs from studies showing a role for caspase 3 in the apoptosis of infected neurons in vitro [38] . Similarly we found no evidence of widespread neurodegeneration in the CNS , although this could be due to the rapid progression of the disease , or result from the absence of IFN responses , which could sensitize the tissue to apoptosis and support CD8+ T cell mediated cytotoxic responses [39] . Alternatively , recent studies suggest that ZIKV infection results in the formation of autophagosomes that facilitate virus replication [40] . Since autophagy is negatively correlated with Type I interferon production [41] , it is possible that the increase in viral load and lack of infiltrating T cells in IFNAR KO mice may be secondary to increased autophagy , reduced antigen clearance by programmed necrosis , and/or reduced presentation . Additional studies will be needed to assess the effects of infiltrating neutrophils and macrophages in the CNS of infected IFNAR KO mice and determine whether they would constitute a therapeutic target in the human disease . Lastly , given the limited neurodegeneration observed in the CNS of IFNAR KO mice at peak of disease and the rapidity of death in the IFNAR KO mice , it is possible that infection of the peripheral organs contribute to the lethality of ZIKV infection in IFNAR KO mice , and thus although the mice show clear signs of neurological damage , additional studies will be needed to establish the cause of death in the infected IFNAR KO mice . The neonatal B6 WT model presents several striking differences with the IFNAR KO model used in these studies . In addition to the differences in the pace ( prodrome of 13 vs 5 days ) and survival , the B6 WT model shows no evidence of virus spread into spleen or liver at the peak of the disease , following subcutaneous infection . This may be secondary to the effect of increased levels of ISG expression resulting in protection of the peripheral organs as well as lower levels of virus in the CNS of B6 WT compared to IFNAR KO mice . In other words , it is conceivable that the absence of virus in the periphery of B6 WT mice reflects the preferential homing of the virus to the CNS or the more efficient IFN-mediated protection of virus in other organs . Indeed the levels of mRNA for type I IFNs at the peak of clinical disease are strikingly low and suggest that the virus may interfere with the IFN response as reported for other viruses [42 , 43] . Additional studies will clarify whether the virus in the B6 WT model exclusively infects the nervous system or whether infections extend to peripheral tissues where it is rapidly cleared and assess whether it extends to the eye as has been reported in the IFNAR KO models . The B6 WT model also differs from the IFNAR KO model in the type of cellular infiltration and immune response in the CNS at the peak of infection , with IFNAR KO displaying extensive infiltration of neutrophils and macrophage as well as upregulation of inflammatory genes such as IL-6 , TNFa and IL-1 , whereas B6 WT mice show cellular infiltrates composed primarily of T cells and upregulation of genes associated with Th1 CD4+T cells and IFN-driven cytotoxic CD8+ T cell responses . Indeed , the differential influx of immune cells evident in the CNS of B6 WT and IFNAR KO mice may underlie the differential clinical outcome . It is possible that the direct damage that Zika causes to the tissue is only one component of its pathogenesis , while the immune response it elicits may also contribute to the pathology . In recent studies we showed that mice infected with Tacaribe virus develop a meningoencephalitis that is driven by the CD4+ and CD8+ T cell response to the virus as mice lacking T cells do not develop disease despite high levels of virus in the CNS[32] . Similar observations were made in mice infected with Sindbis or West Nile virus [44–46] . While it is possible that the influx of T cells into the CNS of B6 WT plays a key role in controlling the virus and aids in the survival of the host , the presence of foci of degenerating neurons suggests that the influx of CD8+ T cells may be driving the observed neuropathology . Understanding the role the different immune cells play in pathogenesis and anti-viral responses will be critical for rational development of therapeutic and preventive approaches . Type I IFNs play a key role in the earliest responses to viral infections and it is well known that several Flaviviruses interfere with IFN production or activity [47 , 48] , however other factors such as age , neuronal and astrocyte maturation are likely important factors for the establishment of productive infections as observed with other neurotropic RNA viruses including Sindbis , Chikungunya and Tacaribe virus , where neonate but not adult mice develop productive infections [25 , 26 , 49] . Similar resistance to challenge was recently reported in studies with ZIKV in IFN deficient mice [28] . In our studies , the virus replicates in neonatal neural tissue in both B6 WT and IFNAR KO mice , but was present in peripheral tissues only in IFNAR KO mice suggesting that: i ) interferons play a role in limiting ZIKV spread but the virus can infect other tissues if the innate immune response to the virus is deficient and ii ) the virus can infect immature nervous tissue even in immunocompetent animals . Detailed studies of the kinetics of infection and clearance in other organs of B6 WT and IFNAR KO mice are underway to understand whether the virus can establish and/or clear productive infections in other immunoprivileged sites . These studies establish and characterize the first contemporary ZIKV animal model in immunocompetent neonatal B6 WT mice . The loss of balance and altered gait observed are consistent with the evidence of neurodegeneration and infiltration by cytotoxic CD8+ T cells in the cerebellum of B6 WT mice . The presence of infiltrating CD8+ and CD4+ T cells also suggests that the virus could induce an immune response that triggers a neurodestructive inflammatory response in the CNS . The B6 WT model will allow for studies into the immunopathology of the virus in a milieu that does not exclude one of the main immune paths for resistance to virus infection . Given the breadth of knock out and transgenic strains available on a C57BL/6 background , it will facilitate detailed investigations into the pathogenesis of the disease as well as mechanistic studies for possible therapeutics . While the need to infect mice at day 1 of life may limit its utility in assessing the direct protective effect of vaccines , it allows for conducting vaccination studies in pregnant mice followed by challenges in the offspring . Lastly , since this model entails a 13 day prodromal phase , it provides an opportunity for the testing of potential therapeutics and its non-lethal outcome allows for studies assessing the long term effects of the infection , and offers the option of testing conditions that may lead to reactivation of the disease .
C57BL/6 ( B6 ) and C57BL/6-IFNAR-/- ( IFNAR KO ) mice used in this study were bred as homozygous breeding pairs ( >20 generations ) . Mice were housed in sterile microisolator cages under 12-hour day/night cycle and given food and water ad libitum in the specific pathogen-free , AAALAC accredited animal facility of the U . S . Food and Drug Administration’s Division of Veterinary Medicine ( Silver Spring , MD ) . This study was carried out in strict accordance with the recommendations in the Public Health Service Policy on Humane care and Use of Laboratory Animals . All protocols involving animals were approved by the Animal Care and Use Committee at US-FDA ( Protocol Number: #2016–14 ) . Zika virus PRVABC59 ( Puerto Rico strain ) used in this study is a contemporary strain that was isolated by CDC from the serum of a ZIKV infected patient who travelled to Puerto Rico in 2015 . The complete genome sequence is published ( Ref . Gene bank accession # KU501215 ) . The virus stocks used for these studies had a titer of 7 . 2 log10 pfu/mL . Virus stocks from CDC were kindly provided by Maria Rios ( Food and Drug Administration ) . All newborn mice were born from pathogen-free parents and inoculated with 2000 PFU or 20 , 000 PFU as indicated by subcutaneous ( s . c . ) inoculation . IFNAR KO mice were inoculated at 10 days of life ( P10 ) and C57BL/6 mice were inoculated one day after birth ( P1 ) . In some experiments IFNAR KO mice were infected on P1 and P3 and C57BL/6 mice were infected on P10 . For experiments tracking survival following ZIKV infection , mice were monitored daily for clinical signs of pathology and weighed every other day to minimize handling . Moribund ( unable to access nutrition due to severe paresis and/or respiratory distress ) animals were euthanized in accordance with the FDA IACUC guidelines . Mice were examined daily for signs of infection and weighed on alternate days . Examination included appearance , stance , and motility . Fig 1C includes a description of the changes observed including the evidence of tremors , hyperactivity ( increased motor exertion and excitability ) , stance ( increased spread of hind legs while standing or walking ) , staggered march ( evidence of unusual pauses during movement ) , limb collapse ( refers to the momentary collapse of the limb under the weight of the body ) , seizures ( partial loss of voluntary movement with evidence of stiffness and/or tonic contraction of the hind legs ) , flaccid paralysis ( loss of muscle tone with collapse of the lower extremities ) . For brain homogenates , infected mice were euthanized by CO2 asphyxiation and exsanguinated by trans-cardiac perfusion . Brains were removed aseptically , placed in 2 ml of cold RPMI media ( ThermoFisher , Carlsbad , CA ) and manually disrupted with ice cold Tenbroeck glass grinders ( Wheaton , Millville , NJ ) until uniform homogenates were obtained . The cellular fractions were pelleted by centrifugation at 400 x g for 15 min . The supernatants were collected and stored at -80°C prior to virus assay . The pelleted ( cellular ) fraction was used for flow cytometry analysis ( see below ) . In some experiments the homogenates were directly stored at -80°C and centrifuged after thawing . The supernatants were then used for the assay . Infectious ZIKV levels were measured as TCID50/0 . 5g of tissue on Vero monolayers using an end-point dilution assay as previously described [50 , 51] . ZIKV RNA levels were measured using quantitative one step reverse transcriptase PCR to amplify ZIKV genome position 1087 to 1163 based on ZIKV MR 766 strain ( GenBank accession no . AY632535 ) Zika virus RNA transcript levels in the samples were quantified by comparing to a standard curve generated using dilutions of an RNA transcript copy of ZIKV sequence . We used 5 micrograms of isolated total RNA for each sample analyzed and ZIKV RNA levels are expressed as ZIKV copies/mL using a standard curve ( S4 Fig ) . The assay can detect all known genotypes of ZIKV and does not cross react with closely related viruses and was performed as described[50] . Briefly , the amplification was performed using 25 ul volume in the Applied Biosystem Viia7 real time PCR machine with the following cycles and conditions: 1st cycle 60C for 30 min followed by 95C for 15 min . The ensuing 45 cycles used 95°C 15 sec and 60°C for 1 min . Brains were collected from infected animals that were exsanguinated by transcardiac perfusion with 10 ml ice-cold PBS . The brains were then bisected along the longitudinal fissure and the entire hemisphere of one half brain was flash frozen in liquid N2 and stored at -80°C . The frozen tissue was homogenized in 2 mL/half-brain of Trizol reagent ( ThermoFisher , Carlsbad , CA ) and RNA was isolated following the manufacturers’ protocol and resuspended in molecular grade ultra-pure ddH2O . The concentration and purity of isolated RNA was determined by spectrophotometry at 260 nm and 280 nm using a NanoDrop 1000 spectrophotometer ( ThermoFisher , Carlsbad , CA ) . To eliminate potential genomic DNA contamination , the DNA-free Turbo kit ( ThermoFisher , Carlsbad , CA ) was used as per the manufacturers protocol . Reverse transcription was performed on 1 μg of total RNA , using Multiscript High Capacity Reverse Transcriptase ( ThermoFisher , Carlsbad , CA ) , per the manufacturer’s protocol , using random primers . The resulting cDNA was diluted ten-fold with ultra-pure water and stored at -20°C prior to use in real-time Taqman PCR reactions ( ThermoFisher , Carlsbad , CA ) . Mouse Immune Array TLDA cards ( ThermoFisher , Carlsbad , CA ) were used as per manufacturers’ instructions . Briefly , the cDNA generated above was diluted 2-fold . Equal volumes of diluted cDNA and 2x Universal Taqman Master Mix was prepared . This mixture was loaded into the chambers of a Taqman Array card , which was centrifuged to distribute the cDNA throughout the card . IFN and interferon stimulated genes ( ISGs ) expression levels were measured using individual Taqman gene assays for IFN-β , ISG-15 , OAS-1g ( ThermoFisher , Carlsbad , CA ) . The cDNA used in the TLDA reactions was also used for these single gene expression assays , with each sample being run in triplicate for each gene . Fold-change in gene expression was determined using the ΔΔCt method [52] , with expression normalized to the expression of the house keeping gene GAPDH . Gene expression is expressed as fold change relative to the indicated uninfected controls . Real-time PCR acquisition and analysis was performed using a Viia7 real-time PCR machine using Quant Studio software , using automatic threshold and endpoint settings ( ThermoFisher , Carlsbad , CA ) . Brains were removed from exsanguinated Zika infected mice or age-matched uninfected control mice , as described above . The brains were then bisected along the longitudinal fissure and one-half of the brain was submerged in 10% neutral buffered formalin ( in phosphate buffered saline ( PBS ) , pH 7 . 4 ) . The other half was flash frozen in liquid N2 for RNA isolation of quantification of virus . After more than 24 hours in fixative , the half-brains were rinsed with PBS and submerged 30% sucrose for cyro-protection . The brains remained in sucrose until they sank in the solution ( typically 24–48 hours ) . Sucrose saturated tissue was then embedded in TissueTek O . C . T ( Sakura-Finetek , Torrance , CA ) by freezing the brains in Tissue-Tek O . C . T . ( Sakura-Finetek , Torrance , CA ) embedding compound using 2 methyl butane cooled with dry ice . The brains were wrapped in aluminum foil and stored at -80°C . Sagittal sections ( 20 μm thick ) were cut using a Leica CM1900 cryostat ( Leica Biosystems , Buffalo Grove , IL ) and thaw-mounted onto SuperFrost-plus microscope slides ( Fisher Scientific , Carlsbad , CA ) . The sections were stored at -80°C until staining . Prior to staining , sections were thawed and dried at room temperature ( RT ) for approximately 10 min , rinsed with phosphate buffered saline ( PBS ) and permeabilized using 0 . 2% Tween-20 in PBS for 20 minutes at RT . For ZIKV staining , antigen unmasking was performed prior to staining by submerging the sections in Sodium citrate solution ( pH8 . 8 , 80°C ) for 30 minute . These sections were then blocked with 1% low fat milk in PBS + 0 . 05% Tween-20 for 1 hour prior to staining . All other sections were blocked with 5% normal goat serum + 1% bovine serum albumin ( BSA ) in PBS with 0 . 05% Tween-20 for at least 60 minutes at RT . Primary antibodies used include: mouse anti-flavivirus mAb ( cloneD1- 4G2-4-15; EMD Millipore ) , human polyclonal anti-zika virus envelope antibody ( Kerafast EVU302 , MA ) , neurofilament heavy chain cocktail [SMI31 and SMI32 mAb] ( Biolegend ) and rat anti-CD45 ( BD Biosciences ) . Tissue sections were incubated overnight in a humidified chamber at 4°C with primary antibodies diluted with 1% BSA in PBS with 0 . 05% Tween-20 . The slides were then rinsed with PBS and incubated with the appropriate AlexaFluor-conjugated ( raised in goat ) secondary antibodies , diluted in 1% BSA in PBS with 0 . 05% Tween-20 ( ThermoFisher , Carlsbad , CA ) for > 60 min at RT . All IF-IHC sections were mounted with ProLong Diamond anti-fade mounting media containing DAPI ( ThermoFisher , Carlsbad , CA ) . Fluorescently labelled antibodies were detected at emission wavelengths: 405 ( DAPI ) , 535 ( Alexa-fluor 488 , Fluoro-Jade C ) , 605 ( Alexa-fluor 568 ) . Sections were imaged using a Pannoramic Digital Slide Scanner . Images were captured using Pannoramic Viewer software ( 3DHistech , Budapest , HUN ) . Fluoro-Jade C histology staining , which specifically detects neuronal degeneration , [53] was performed as per the manufacturers protocol ( EMD Millipore , Billerica , MA ) . For confocal imaging , sections were imaged using a Zeiss LSM 880 confocal microscope , using a 405 nm , 488 nm , and 561 nm excitation lasers . Optimal fluorescence detection settings were determined using B6 WT sections and applied to all other sections . Images were acquired using the Zeiss Zen software using a Z-stack , with slices of 0 . 44–0 . 49 microns per slice for all sections . Laser and PMT setting were consistent for uninfected and infected sections . Maximum intensity projections of these Z-stacks were generated using the Zen software . All images were prepared for publication using Adobe Photoshop CC 2015 software . The cellular fractions of brain homogenates ( see Virus Quantification ) were pooled and resuspended in 30% percoll ( GE Life Sciences , Marlborough , MA ) in RPMI + 25 mM HEPES ( ThermoFisher , Carlsbad , CA ) and underlayed with 1 ml of 70% percoll . After centrifugation at 800 x g for 30 minutes , CNS cells were collected from the 30%-70% interface , washed in RPMI and isolated by centrifugation at 400 x g for 10 minutes . Non-specific antibody binding was blocked with a mix of mouse Fc block ( purified α-CD16/32 , BD Biosciences ) and normal mouse serum for at least 15 minutes . All antibodies ( anti-CD45 , anti-CD4 , anti-CD8 , anti-CD11b , anti-CD19 anti-Ly6G , anti-NK1 . 1 ( BD Biosciences , San Jose , CA ) ) ; and anti-F4/80 ( AbD Serotec , Raleigh , NC ) were directly conjugated with one of the following fluorochromes: fluorescein isothiocyanate ( FITC ) , phycoerythrin ( PE ) , peridinin-chlorophyll proteins ( PerCP ) , or alophycocyanin ( APC ) . The cells were incubated with these antibodies for 20 min in FACS buffer ( 1% BSA in PBS ) , then washed with FACS buffer , fixed in 2% paraformaldehyde solution and acquired using a BD Fortessa flow cytometer ( BD Biosciences , San Jose , CA ) . The flow cytometer was calibrated using beads ( eBiosciences ) conjugated with fluorescent antibodies for each channel in the assay . The resulting data from each samples was analysed using FlowJo ( version 10 ) software ( FlowJo , LLC , Ashland , OR ) .
|
The recent spread of Zika virus ( ZIKV ) and its association with increased rates of neurological disorders and congenital defects created an urgent need for animal models to examine the pathogenesis of the disease and explore the efficacy of potential therapeutics and vaccines . We describe the first symptomatic PRVABC59 ( ZIKV ) animal model in immunocompetent B6 WT mice showing that a subcutaneous challenge in 1 day old mice leads to non-lethal neurological disease that is characterized by unsteady gait , kinetic tremors , severe ataxia and seizures that subsides after 2 weeks . ZIKV infects neurons in cerebellum of mice and elicits the infiltration of lymphocytes into the brain . The immune response protects mice from death but may also contribute to neurodegeneration as mice with defective interferon responses have increased virus loads in brain and peripheral organs , succumbing to the disease in 5–6 days , but have fewer signs of neurodegeneration . This mouse model bypasses transplacental transmission and consequent placental insufficiency and will facilitate detailed investigations into the pathogenesis of the disease as well as mechanistic studies for possible therapeutics and vaccines . Lastly , its non-lethal outcome allows for studies assessing the long term effects of the infection , and exploring conditions that could lead to disease reactivation .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"pathology",
"and",
"laboratory",
"medicine",
"nervous",
"system",
"pathogens",
"immunology",
"microbiology",
"brain",
"animal",
"models",
"viruses",
"model",
"organisms",
"rna",
"viruses",
"signs",
"and",
"symptoms",
"cerebellum",
"research",
"and",
"analysis",
"methods",
"white",
"blood",
"cells",
"inflammation",
"animal",
"cells",
"proteins",
"medical",
"microbiology",
"microbial",
"pathogens",
"t",
"cells",
"mouse",
"models",
"immune",
"response",
"biochemistry",
"diagnostic",
"medicine",
"anatomy",
"central",
"nervous",
"system",
"flaviviruses",
"cell",
"biology",
"viral",
"pathogens",
"interferons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"cerebral",
"cortex",
"organisms",
"zika",
"virus"
] |
2016
|
Zika (PRVABC59) Infection Is Associated with T cell Infiltration and Neurodegeneration in CNS of Immunocompetent Neonatal C57Bl/6 Mice
|
It has been suggested that some cancer cells rely upon fatty acid oxidation ( FAO ) for energy . Here we show that when FAO was reduced approximately 90% by pharmacological inhibition of carnitine palmitoyltransferase I ( CPT1 ) with low concentrations of etomoxir , the proliferation rate of various cancer cells was unaffected . Efforts to pharmacologically inhibit FAO more than 90% revealed that high concentrations of etomoxir ( 200 μM ) have an off-target effect of inhibiting complex I of the electron transport chain . Surprisingly , however , when FAO was reduced further by genetic knockdown of CPT1 , the proliferation rate of these same cells decreased nearly 2-fold and could not be restored by acetate or octanoic acid supplementation . Moreover , CPT1 knockdowns had altered mitochondrial morphology and impaired mitochondrial coupling , whereas cells in which CPT1 had been approximately 90% inhibited by etomoxir did not . Lipidomic profiling of mitochondria isolated from CPT1 knockdowns showed depleted concentrations of complex structural and signaling lipids . Additionally , expression of a catalytically dead CPT1 in CPT1 knockdowns did not restore mitochondrial coupling . Taken together , these results suggest that transport of at least some long-chain fatty acids into the mitochondria by CPT1 may be required for anabolic processes that support healthy mitochondrial function and cancer cell proliferation independent of FAO .
During the last decade , carnitine palmitoyltransferase I ( CPT1 ) has been identified as a potential therapeutic target for a growing list of cancers that include breast cancer , prostate cancer , glioblastoma , colon cancer , gastric cancer , myeloma , and others [1–6] . In these cancers , CPT1 expression is increased , and/or CPT1 inhibition is reported to have antitumor effects . CPT1 is an enzyme associated with the outer mitochondrial membrane that transfers a long chain acyl group from coenzyme A to carnitine [7 , 8] . Importantly , this transformation is required to transport long-chain fatty acids into the mitochondrial matrix . Long-chain fatty acids reaching the mitochondrial matrix are generally assumed to be oxidatively degraded , thereby implicating fatty acid oxidation ( FAO ) as a potentially important pathway in cancer metabolism [9] . FAO is thought to support cancer metabolism primarily in 2 ways . First , given their highly reduced state , fatty acids may provide an important source of ATP to fuel tumor growth [10] . For every pair of carbons in a fatty acid that is completely oxidized , up to 14 ATP can be produced—assuming NADH and FADH2 yield 2 . 5 and 1 . 5 ATP , respectively [11] . Ten of these 14 ATP are produced by oxidizing acetyl-CoA in the tricarboxylic acid ( TCA ) cycle . Oxidation of exogenous fatty acids might be particularly relevant to tumors that grow in adipocyte-rich environments , such as breast cancer [12] . Here , fatty acids transported from neighboring adipocytes may constitute an important energy reservoir [13] . A second potential benefit of cancer cells oxidizing fatty acids is the production of NADPH [14] . Although FAO does not make NADPH directly , the acetyl-CoA it produces in the mitochondria can be shuttled to the cytosol as citrate once acetyl-CoA condenses with oxaloacetate . Each molecule of citrate exported to the cytosol can then produce 1 molecule of NADPH through either isocitrate dehydrogenase 1 or malic enzyme 1 . It has been suggested that some cancer cells rely on this source of NADPH to neutralize oxidative stress [9] . Indeed , inhibition of CPT1 in human glioblastoma cells causes a reduction in NADPH levels and an increase in reactive oxygen species [15] . A major challenge of considering FAO as an essential pathway in cancer metabolism is that cancer cells are also thought to rely heavily on fatty acid synthesis [16] . While one can rationalize the coexistence of FAO and fatty acid synthesis on the basis of subcellular compartmentalization , conventional thinking would indicate that it is unproductive to run both pathways simultaneously [9] . Additionally , recent data from our laboratory suggest that such a futile cycling process occurs to only a minimal extent in at least some proliferating cells [17] . As noted , the focus on FAO in cancer cells has mostly been driven by experimental findings related to CPT1 [6] . The assumption has been that increased CPT1 expression and sensitivity to CPT1 inhibition represents a demand for FAO . In this work , we consider an alternative possibility that CPT1 has important metabolic roles independent of FAO . We present evidence that long-chain fatty acids transported into the mitochondria via CPT1 have important anabolic fates that are essential for proliferation . We also provide data suggesting that etomoxir , a drug commonly used to inhibit CPT1 in cancer studies , has off-target effects that may complicate the interpretation of some experiments . We focus much of our attention on the breast cancer cell line BT549 , because the essential role of CPT1 in these cells has already been thoroughly demonstrated [18] . We show that inhibiting FAO by as much as 90% had no effect on BT549 cell proliferation . At this level of pharmacological CPT1 inhibition , minimal labeling from 13C-enriched fatty acids could be detected in citrate . These results suggest that BT549 cells do not require FAO as a major source of ATP or NADPH . When CPT1 is knocked down , however , we found that BT549 cell proliferation was significantly reduced . Under these conditions , the function of the mitochondria was impaired , and changes in the levels of complex lipids within mitochondria were detected . The cells could not be rescued by acetate or octanoic acid supplementation . These data support a role for CPT1 in the proliferation of some cancer cells that is independent of FAO .
All cells were cultured in high-glucose DMEM ( Life Technologies ) containing 10% FBS ( Life Technologies ) and 1% penicillin/streptomycin ( Life Technologies ) at 37 °C with 5% CO2 . All culture media for growing cells were supplemented with 100 μM palmitate-BSA and 100 μM oleate-BSA to approach the physiological concentrations of free fatty acids . When counting cells manually , BT549 cell media were refreshed to control or experimental media 24 hours after the cells were seeded ( at t = 0 ) to assess growth . At selected time points , cells were collected and counted in trypan blue with an automatic cell counter ( Nexcelom ) . Doubling time was calculated by linear regression against the logarithm of cell density in exponential phase . For assessing proliferation , cells were grown under various experimental conditions for 48 hours , and proliferation was determined by using an MTT assay ( ATCC ) according to the manufacturer’s instructions . Absorbance was measured at 570 nm by using the Cytation 5 microplate reader ( BioTek ) with a reference wavelength set at 670 nm . We note that comparable changes in cell proliferation were measured using the MTT assay and manual cell counting when BT549 cells were treated with 200 μM etomoxir for 48 hours ( S1 Fig ) , indicating that the 2 techniques to assess cell proliferation provided consistent results in our experiments . Etomoxir was purchased from Cayman Chemical ( purity ≥ 98% ) . Etomoxir was dissolved in water to create a concentrated stock solution . The vehicle control was water alone . CPT1A silencing was achieved by using a validated pool of small interfering RNA ( siRNA ) duplexes directed against human CPT1A ( Trifekta Kit , IDT ) and Lipofectamine RNAiMAX Transfection Reagent ( Invitrogen ) according to the manufacturer’s instructions ( see S1 Text for the dicer-substrate short interfering RNA [DsiRNA] sequence ) [19] . The knockdown ( KD ) efficiency was determined by measuring CPT1A mRNA levels with a premade primer ( IDT ) and quantitative RT-PCR ( Applied Biosystems ) . The expression levels were normalized to an HPRT endogenous control . Cells given scrambled siRNA were used as a negative control . For overexpression of human CPT1A , the cDNA was cloned in the pcDNA3 . 1+ vector ( GenScript ) under a constitutive CMV promoter . The codon was optimized to be resistant to the siRNA added . The catalytically dead CPT1A had an identical sequence ( see S1 Text ) to the wild-type siRNA-resistant CPT1A , with the exception of G709E and G710E mutations to abolish catalytic activity ( GenScript ) . For transduction , CPT1A was first knocked down with Lipofectamine RNAiMAX for 24 hours . Next , cells were transduced with plasmids using Lipofectamine 3000 ( Invitrogen ) for 4 hours . Media were then refreshed , and cells were assayed 48 hours post plasmid transduction ( 72 hours post siRNA knockdown ) . The control vector was pcDNA3 . 1+N-eGFP ( GenScript ) , which expresses GFP instead of CPT1A . BT549 cells were treated with either a scrambled siRNA control or siRNA targeting CPT1A for 12 hours . Next , nutrients were added to each culture plate and incubated for 48 hours before assessing cell proliferation with an MTT assay . Each compound ( sodium acetate , octanoic acid , uridine , and sodium pyruvate ) was added separately and evaluated in an independent experiment relative to vehicle controls . For sodium acetate , the vehicle control was sodium chloride . Cells were lysed with RIPA buffer ( Thermo Fisher Scientific ) in the presence of a protease inhibitor cocktail ( Thermo Fisher Scientific ) and sonicated for 30 seconds . Lysates were separated by SDS-PAGE under reducing conditions , transferred to a PVDF membrane , and analyzed by immunoblotting . Rabbit anti-CPT1A ( No . 12252 ) ( Cell Signaling Technology ) was used as a primary antibody . Immunoblotting for β-tubulin by mouse anti-β-tubulin antibody ( Santa Cruz Biotechnology ) and COX IV by rabbit anti-COX IV antibody ( Cell Signaling ) was used as a loading control for whole-cell lysates and mitochondrial lysates , respectively . Anti-rabbit and anti-mouse secondary antibodies were from Cell Signaling Technology and Thermo Fisher Scientific , respectively . Signal was detected using the ECL system with X-ray film development ( Thermo Fisher Scientific and GE Healthcare Life Sciences ) or a LI-COR C-Digit blot scanner ( LI-COR ) according to the manufacturer’s instructions . Cells were preincubated with the vehicle control or 200 μM etomoxir for 48 hours . On the day of the assay , cells were trypsinized , washed 2 times with cold PBS buffer , and extracted according to the manufacturer’s instructions . The NADH/NAD+ ratio was measured and calculated using an NAD/NADH Quantification Colorimetric Kit ( BioVision ) . To assess the activity of FAO , cells were treated with vehicle control , etomoxir , scrambled siRNA , or CPT1A siRNA for 48–72 hours . Next , the medium was refreshed with new medium containing 100 μM uniformly 13C labeled ( U-13C ) palmitate-BSA and 100 μM natural abundance oleate-BSA . After labeling for 24 hours , cells were harvested , extracted , and analyzed as previously described [17] . For U-13C glucose , U-13C glutamine , and U-13C palmitate tracing experiments , cells were transferred to media containing 13C label and either vehicle control or 200 μM etomoxir for 12 hours , 6 hours , and 24 hours , respectively . The polar portion of the extract was separated by using a Luna aminopropyl column ( 3 μm , 150 mm × 1 . 0 mm I . D . , Phenomenex ) coupled to an Agilent 1260 capillary HPLC system . Mass spectrometry detection was carried out on an Agilent 6540 Q-TOF coupled with an ESI source operated in negative mode . Isotopic labeling was assessed comprehensively by using the X13CMS software [20] . The identity of each metabolite was confirmed by matching retention times and MS/MS fragmentation data to standard compounds . The isotopologue distribution patterns presented were obtained from manual evaluation of the data and calculated by normalizing the sum of all isotopologues to 100% . Data presented were corrected for natural abundance and isotope impurity . After incubating cells in fresh media for 24 hours , the spent media were collected and analyzed . Known concentrations of U-13C internal standards ( glucose , lactate , glutamine , glutamate , and palmitate; Cambridge Isotopes ) were spiked into media samples before extraction . Extractions were performed in glass to avoid plastic contamination as previously reported [21] . Samples were measured by LC/MS analysis , with the method described above . For each compound , the absolute concentrations were determined by calculating the ratio between the fully unlabeled peak from samples and the fully labeled peak from standards . The consumption rates were normalized by cell growth over the experimental time period . Mitochondria were isolated from BT549 cells as previously described [22] . In brief , cells were harvested , pelleted , and resuspended in cold mitochondrial isolation medium ( MIM ) ( 300 mM sucrose , 10 mM sodium 4- ( 2-hydroxyethyl ) -1-piperazineethanesulfonic acid [HEPES] , 0 . 2 mM ethylenediaminetetraacetic acid [EDTA] , and 1 mg/mL bovine serum albumin [BSA] , pH 7 . 4 ) . Cells were then homogenized with a glass-Teflon potter . After homogenization , samples were centrifuged at 700 g at 4 °C for 7 minutes . The supernatant containing mitochondria was centrifuged again at 10 , 000 g for 10 minutes . Mitochondrial pellets were washed with cold BSA-free MIM , and the protein amount was determined by using a Bradford protein assay ( Bio-Rad ) . The oxygen consumption rate ( OCR ) of whole cells and isolated mitochondria was determined by using the Seahorse XFp Extracellular Flux Analyzer ( Seahorse Bioscience ) . Cells were first incubated with vehicle control , 10 μM etomoxir , or 200 μM etomoxir for 1 hour prior to measuring respiration ( we note that etomoxir was present in the assay medium as well ) . For CPT1A knockdowns , cells were treated with scrambled siRNA control or CPT1A siRNA for 48 hours . Cells were trypsinized and plated on a miniplate with the same seeding density 24 hours prior to the Seahorse assay . The assay medium consisted of 25 mM glucose , 4 mM glutamine , 100 μM palmitate-BSA , and 100 μM oleate-BSA in Seahorse base medium . The OCR was monitored upon serial injections of oligomycin ( oligo , 2 μM ) , FCCP ( 1 μM , optimized ) , and a rotenone/antimycin A mixture ( rot/AA , 1 μM ) . To measure the respiration of isolated mitochondria , freshly isolated mitochondria from BT549 cells were resuspended in cold mitochondrial assay solution ( MAS ) . For the composition of MAS , see [22] . Samples were loaded on a miniplate with 5 μg of protein per well . Mitochondria were attached to the plate by centrifuging at 2 , 000 g ( 4 °C ) for 20 minutes . After centrifugation , prewarmed MAS-containing substrates ( 10 mM pyruvate , 2 mM malate , 4 mM adenosine diphosphate ( ADP ) , vehicle control , or etomoxir ) were added to each well without disturbing the mitochondrial layer and then inserted into the XFp analyzer . OCR was monitored upon serial injections of rotenone ( rot , 2 μM ) , succinate ( suc , 10 mM ) , and antimycin A ( AA , 4 μM ) . Whole-cell OCR was normalized to the final cell number as determined by manual cell counting . Data presented were corrected for nonmitochondrial respiration . Cells were incubated with 100 nM MitoTracker Red CMXRos ( Thermo Fisher Scientific ) or 4 μM JC-1 ( Cayman Chemical ) dissolved in complete media at 37 °C for 30 minutes . Cells were washed twice with PBS and then subjected to live imaging , or cells were fixed with 4% paraformaldehyde in PBS . Fixed cells were permeabilized with 0 . 1% Triton X-100 ( Sigma Aldrich ) . Next , cells were washed twice with PBS , and nuclei were stained with DAPI . Cells were then mounted with ProLong Gold ( Thermo Fisher Scientific ) . For live imaging , nuclei were stained with Hoechst 33342 ( Thermo Fisher Scientific ) . Cells were imaged using a Zeiss LSM 880 confocal microscope equipped with Airyscan . Images were acquired with a Zeiss 20x , 40x , 63x/1 . 4 NA objective using the ZEN Black acquisition software . Samples were excited with 405 ( for DAPI and Hoechst 33342 ) , 514 ( for JC-1 monomers ) , and 543 ( for Mitotracker Red and JC-1 aggregates ) laser lines . Images were processed and prepared using the ZEN Black software . Samples were fixed in 2% paraformaldehyde/2 . 5% glutaraldehyde ( Polysciences ) in 100 mM sodium cacodylate buffer , pH 7 . 2 , for 1 hour at room temperature . Samples were washed in sodium cacodylate buffer and postfixed in 1% osmium tetroxide ( Polysciences ) for 1 hour . Next , samples were rinsed extensively in dH2O prior to en bloc staining with 1% aqueous uranyl acetate ( Ted Pella ) for 1 hour . Following several rinses in dH2O , samples were dehydrated in a graded series of ethanol and embedded in Eponate 12 resin ( Ted Pella ) . Sections of 95 nm were cut with a Leica Ultracut UCT ultramicrotome ( Leica Microsystems ) , stained with uranyl acetate and lead citrate , and viewed on a JEOL 1200 EX transmission electron microscope ( JEOL USA ) equipped with an AMT 8 megapixel digital camera and AMT Image Capture Engine V602 software ( Advanced Microscopy Techniques ) . Isolated mitochondria ( with known concentrations of internal standards ) were extracted with chloroform/methanol/water ( 1:1:1 ) and vortexed for 1 minute . After centrifuging at 3 , 000 g for 15 minutes , the chloroform layer was dried under nitrogen gas and reconstituted with methanol/chloroform ( 95:5 ) according to the protein amount . Samples were separated using a Kinetex evo C18 column ( 2 . 6 um , 150 mm × 2 . 0 mm I . D . , Phenomenex ) coupled to an Agilent 1290 UPLC system . Mass spectrometry detection was carried out on an Agilent 6540 Q-TOF or a Thermo Scientific Q Exactive Plus coupled with an ESI source operated in both negative mode and positive mode . The lipid identities were confirmed by accurate mass as well as by matching retention times and MS/MS fragmentation patterns to standards . Absolute quantitation was achieved by normalizing to internal standards for ( PC ( 14:1/14:1 ) , PE ( 16:1/16:1 ) , CL ( 14:0/14:0/14:0/14:0 ) , PG ( 15:0/15:0 ) , PS ( 14:0/14:0 ) , PA ( 12:0/12:0 ) , LPE ( 14:0 ) , LPC ( 17:0 ) , SM ( d18:1/12:0 ) , and Cer ( d18:1/17:0 ) ) .
The first question we sought to address is whether FAO is dispensable in rapidly proliferating cancer cells , such as BT549 . We pharmacologically targeted FAO by using the drug etomoxir ( ethyl 2-[6- ( 4-chlorophenoxy ) hexyl]oxirane-2-carboxylate ) , which has been regarded as a specific inhibitor of CPT1 [23 , 24] . It is common in cancer studies to use etomoxir at hundreds of micromolar concentrations [5 , 15 , 18 , 25 , 26] . Here , we started by considering etomoxir at doses an order of magnitude lower . When BT549 cells were treated with 10 μM etomoxir , we measured over an 80% decrease in acylcarnitine species ( the products of CPT1 activity , Fig 1A ) . Since changes in acylcarnitine levels may not reflect the same change in FAO , we directly assessed FAO by feeding cells uniformly labeled 13C-palmitate ( U-13C palmitate ) and measuring the labeling of FAO products . During FAO , U-13C palmitate is degraded to 13C2-acetyl-CoA . This acetyl-CoA then condenses with oxaloacetate in the TCA cycle to produce 13C2-citrate ( the M+2 isotopologue ) . Upon treatment with 10 μM etomoxir , 13C2-citrate labeling from U-13C palmitate decreased by approximately 90% compared to vehicle controls ( Fig 1B ) . These data demonstrate that 10 μM of etomoxir effectively blocks most of FAO . Surprisingly , 10 μM etomoxir did not affect the proliferation rate of BT549 cells relative to vehicle controls ( Fig 1C ) . Increasing the concentration of etomoxir by a factor of 10 to 100 μM led to further decreases in acylcarnitine levels and citrate labeling from U-13C palmitate , but we still did not observe a statistically significant change in the proliferation rate of BT549 cells ( S2A Fig ) . Comparable results were observed in HeLa cells . When HeLa cells were treated with 100 μM etomoxir , no FAO activity was detected , yet we observed no alteration in proliferation ( S2B and S2C Fig ) . An analysis of 6 additional cell lines produced similar results for B16 , 3T3 , MCF7 , and HS578t cells ( S2A Fig ) . Only 2 cell lines tested ( H460 and T47D ) showed a statistically significant decrease in proliferation with 100 μM etomoxir treatment . These data suggest that FAO is not an essential source of ATP or NADPH in some cancer cells , such as BT549 . Given that studies evaluating the role of CPT1 in cancer have commonly used concentrations of etomoxir at the hundreds of micromolar or even 1 mM level [15] , we next assessed whether higher concentrations of etomoxir affected cell growth . Although 10 μM etomoxir was sufficient to inhibit most of FAO , residual FAO could be further reduced with increasing concentrations of etomoxir . This was reflected by additional small decreases in acylcarnitine pools ( Fig 1A ) and additional small decreases in the labeling of citrate from U-13C palmitate ( Fig 1B ) . Despite the relatively small differences in FAO between 10 and 200 μM etomoxir-treated BT549 cells , we found that 200 μM of etomoxir resulted in a statistically significant reduction in cellular proliferation rate , while 10 μM did not ( Fig 1C ) . These data are consistent with previous reports of the effects of 200 μM etomoxir on BT549 cells [18] . Interestingly , even though no FAO could be measured at 200 μM ( Fig 1B ) , higher concentrations of etomoxir continued to result in further reductions in cell proliferation for BT549 cells ( S3A Fig ) . Similar results were obtained from other cell lines tested ( S3B Fig ) . Taken together , these observations suggest that high concentrations of etomoxir influence proliferation rate independent of FAO . Since impairing approximately 90% of FAO did not change the rate of BT549 cell proliferation , we hypothesized that these cells might compensate for losses in ATP or NADPH production by increasing the oxidation of metabolic substrates other than fatty acids ( e . g . , glucose or glutamine ) . We therefore analyzed cell culture media to evaluate nutrient-uptake and waste-excretion rates of cells treated with etomoxir . Interestingly , when 90% of FAO was inhibited with 10 μM etomoxir , we observed no change in the rate of glucose uptake or lactate excretion ( Fig 2A ) . Instead , with 10 μM etomoxir , we observed a 30% decrease in glutamate excretion . We note that cells treated with 10 μM etomoxir did not alter their glutamine uptake . These data suggest that when FAO is mostly blocked , BT549 cells can possibly compensate for the loss of energy/reducing equivalents by up-regulating glutaminolysis , by which glutamine carbons are fed into the TCA cycle instead of being excreted as glutamate ( see S11 Fig , introduced below ) . Additionally , we observed a 30% decrease in the uptake rate of fatty acids ( palmitate and oleate ) in etomoxir-treated cells compared to vehicle controls , presumably because drug-treated cells cannot degrade these fatty acids by FAO . Consistent with our proliferation results , different concentrations of etomoxir resulted in strikingly distinct nutrient utilization profiles that did not correlate with the small differences we observed in FAO . While 10 μM etomoxir did not change glucose uptake or lactate excretion , we observed an increase in glycolysis ( indicated by a 3 . 5% increase in glucose consumption and a 4 . 8% increase in lactate excretion ) when cells were given 200 μM etomoxir ( Fig 2A ) . We also found that in contrast to the decrease in palmitate and oleate uptake we observed in cells treated with 10 μM etomoxir , cells treated with 200 μM etomoxir took up approximately 30% more palmitate and oleate even though these fatty acids could not be oxidized . Most notably , instead of decreasing by 30% as we observed with 10 μM etomoxir treatment , glutamate excretion increased by nearly 2-fold with 200 μM etomoxir ( Fig 2A ) . Considering that glutamine uptake was unaltered , this result suggests that less glutamine carbon is available for oxidation in the TCA cycle at high concentrations of etomoxir ( see S11 Fig , introduced below ) . Although etomoxir is often assumed to be a specific inhibitor of CPT1 , our observations above prompted us to consider other possible off-target activities , particularly at high drug concentrations as are often used in cancer studies [27] . We first examined the OCRs of BT549 cells treated with etomoxir . Cells were assayed in nutrient-rich media containing 25 mM glucose , 4 mM glutamine , 100 μM palmitate , and 100 μM oleate in the presence of vehicle control or etomoxir . Cells were treated with etomoxir for 60 minutes prior to making the oxygen consumption measurements . As we expected based on our nutrient-utilization data and proliferation results , the mitochondrial respiration profiles of cells treated with 10 μM etomoxir were not significantly different from vehicle controls ( Fig 2B ) . With 200 μM etomoxir , however , mitochondrial respiration was significantly impaired . We measured a 65% decrease in basal respiration and a 65% decrease in maximal respiratory capacity after treating cells with 200 μM etomoxir . Moreover , we detected only minimal oxygen-driven ATP production , and the calculated mitochondrial coupling efficiency was therefore determined to be nearly zero ( Fig 2C ) . Given the impaired mitochondrial respiration that we observed with 200 μM etomoxir treatment in whole cells , we hypothesized that high concentrations of etomoxir might directly inhibit the activity of the electron transport chain . To test this possibility , we isolated intact mitochondria from BT549 cells and measured changes in oxygen consumption upon etomoxir treatment . By using isolated mitochondria instead of whole cells , we could control the availability of substrates for respiration . For this experiment , it is critical to point out that isolated mitochondria were assayed in buffer free of fatty acids , acyl-CoA species , acylcarnitines , and carnitine . Under such conditions , no FAO is occurring , and hence , CPT1 inhibition will not affect respiration . Any change in oxygen consumption upon etomoxir administration can therefore be attributed to off-target effects . We evaluated mitochondrial respiration in 4 time segments over which various respiratory substrates and inhibitors were added ( Fig 2D ) . The purpose of this experimental design was to distinguish respiration driven by complex I ( state I respiration ) from respiration driven by complex II ( state II respiration ) . At time zero , mitochondria were provided pyruvate , malate , and ADP . These substrates enable turnover of the TCA cycle and production of NADH . Oxidation of NADH by respiratory complex I drives oxygen consumption . In time segment 2 , we added rotenone to the mitochondria . Rotenone inhibits complex I and therefore blocks oxygen consumption under these conditions by preventing the electron transport chain from accepting its only source of electrons . In time segment 3 , we provided mitochondria an alternative source of electrons in the substrate succinate . Oxidation of succinate feeds electrons into respiratory complex II of the electron transport chain , which is independent of the rotenone-inhibited complex I and therefore reintroduces oxygen consumption . Finally , in time segment 4 , mitochondria were treated with antimycin A . Antimycin A inhibits respiratory complex III , thereby preventing the electron transport chain from oxidizing any of the substrates present . Under these conditions , there is no mitochondrial oxygen consumption . Data from vehicle controls ( Fig 2D , black ) were as expected . Next , we independently considered isolated BT549 mitochondria treated with 200 μM etomoxir for 15 minutes . We performed the respiration measurements detailed above over the same 4 time segments . Notably , relative to the vehicle controls , there was a 35% decrease in state I respiration upon etomoxir treatment ( Fig 2D , red ) . However , there was no statistically significant change in state II respiration . In contrast , 10 μM etomoxir resulted in similar OCRs for both state I and state II respiration ( S4A Fig ) . These data suggest that high concentrations ( 200 μM ) of etomoxir inhibit respiratory complex I but do not affect downstream proteins in the electron transport chain . The results also indicate that low concentrations ( 10 μM ) of etomoxir do not have off-target effects on the electron transport chain ( S4A Fig ) . Similar to etomoxir , we note that the complex I inhibitor rotenone also slowed down BT549 cell proliferation when given in culture media ( S4B Fig ) . We surmised that this off-target effect of 200 μM etomoxir on respiratory complex I might prevent regeneration of NAD+ from NADH and hence inhibit the turnover of the TCA cycle , thereby contributing to increased glycolysis and decreased glutaminolysis . Indeed , the intracellular NADH/NAD+ ratio was increased in cells treated with 200 μM etomoxir ( S5 Fig ) . To further test our prediction , we fed BT549 cells U-13C glucose or U-13C glutamine and measured labeling in TCA cycle metabolites . Compared to vehicle controls , labeling of glycolytic intermediates from U-13C glucose was slightly increased , while labeling of TCA cycle metabolites from U-13C glucose was slightly decreased in cells treated with 200 μM etomoxir ( Fig 2E , S6 and S7 Figs ) . These data are consistent with results shown in Fig 2A , indicating that cells treated with 200 μM etomoxir direct more glucose carbon into lactate instead of aerobic respiration . Although BT549 cells treated with 200 μM etomoxir showed only a modest increase in glycolysis , we note that much larger increases in glycolysis were observed for other cell lines treated with 200 μM etomoxir ( S8 Fig ) . In BT549 cells treated with 200 μM etomoxir , we also observed a decrease in the overall labeling of citrate and other TCA cycle intermediates from U-13C glutamine relative to vehicle controls ( Fig 2F , S9 Fig ) . Additionally , the pools of TCA cycle intermediates were decreased , with the exception of α-ketoglutarate , which is the entry point of glutamine into the TCA cycle ( S10 Fig ) . These results are consistent with decreased glutaminolysis , indicated by similar glutamine uptake but increased glutamate excretion ( Fig 2A ) . The relative TCA cycle activity can also be inferred by the ratio of the M+2 isotopologue to the M+4 isotopologue ( i . e . , M+2/M+4 ) of malate ( S11A Fig ) . The M+2/M+4 ratio was higher when BT549 cells were treated with 10 μM etomoxir compared to vehicle control , while the M+2/M+4 ratio was lower when BT549 cells were treated with 200 μM etomoxir ( S11B Fig ) . Interestingly , in cells treated with 200 μM etomoxir , we detected increased labeling of the M+5 isotopologue in citrate from U-13C glutamine . This result is consistent with a relative increase in the reductive metabolism of glutamine , which is a metabolic signature of cells under hypoxic stress [28] . Having established that etomoxir has off-target effects , we chose to use genetic methods to inactivate CPT1 . There are 3 subtypes of CPT1 that are encoded by different genes and show tissue-specific distribution [29] . CPT1B is expressed in muscle , heart , and adipose tissue and CPT1C in neurons , whereas CPT1A is more widely expressed and has been previously implicated as a therapeutic target in breast cancer cells [2 , 30 , 31] . Using siRNA , we knocked down CPT1A mRNA levels by >90% relative to scrambled siRNA controls ( S12A Fig ) . All of our assays to phenotype CPT1A knockdown ( CPT1AKD ) cells were performed at least 48 hours post transfection and completed within 96 hours , over which time CPT1A mRNA levels and protein levels remained greatly reduced ( S12B Fig ) . As evidence that knockdown of CPT1A blocked transport of fatty acids into the mitochondria , we observed major reductions in the levels of acylcarnitine species ( S13 Fig ) , and we detected no 13C-labeled citrate after 24 hours of U-13C palmitate labeling ( Fig 3A ) . These data indicated that CPT1A knockdown inactivated most of FAO . Notably , CPT1AKD cells had a significantly impaired proliferation rate ( Fig 3B ) , with a 50% increase in doubling time ( 42 . 5 hours ) compared to control wild-type cells with scrambled siRNA ( 27 . 8 hours ) . Given that the end product of β-oxidation is acetyl-CoA and that acetyl-CoA is readily produced from acetate , acetate supplementation has been shown to rescue cellular functions dependent upon FAO [19] . In our cells , however , impaired proliferation due to CPT1A knockdown could not be rescued by acetate supplementation ( Fig 3C ) , suggesting again that CPT1A affects the growth of BT549 cells independent of FAO . Interestingly , supplementation of acetate slightly impaired BT549 cell growth . This could be partially explained by the osmotic effects of sodium , acetate’s counter ion ( S14 Fig ) . We also attempted to rescue the proliferation of knockdown cells by supplementing them with octanoic acid , which can passively diffuse through the inner mitochondrial membrane independent of CPT1 and therefore compensate for impaired FAO [32] . Similar to acetate , supplementing cells with various concentrations of octanoic acid did not restore their proliferation ( S15 Fig ) , further supporting that CPT1A knockdown influences cell phenotype independent of FAO . To rule out the possibility that decreased cell proliferation in CPT1A knockdown cells was a result of off-target effects of siRNA , we performed 2 analyses . First , we tested 2 different siRNA sequences and observed comparable protein depletion and growth inhibition in both ( S16A and S16B Fig ) . Given that growth inhibition is a common off-target effect of siRNA , however , we performed a second experiment in which we attempted to rescue CPT1A knockdown cells by overexpressing siRNA-resistant CPT1A protein ( CPT1Aresistant ) ( S16C Fig ) . CPT1Aresistant protein led to a significant increase in FAO and cellular proliferation rate relative to vector controls ( S16D and S16E Fig ) . Together , these data indicate that decreased proliferation in siRNA-treated cells is due to CPT1A loss of function rather than off-target effects . We also observed changes in nutrient utilization upon CPT1A knockdown ( Fig 4A ) . CPT1AKD cells had a nearly 2-fold increase in glucose uptake and lactate production relative to scrambled siRNA controls , indicating a substantial increase in glycolytic flux . Additionally , relative to wild-type cells with scrambled siRNA , CPT1AKD cells had a 2-fold increase in palmitate uptake and a 6 . 5-fold increase in oleate uptake . Yet , in contrast to cells treated with 200 μM etomoxir , CPT1AKD cells increased their uptake of glutamine by 45% and began uptaking glutamate instead of excreting it ( Fig 4A ) . The increased utilization of glutamine and glutamate carbon suggests increased glutaminolysis and thus increased TCA cycle activity in CPT1AKD cells , whereas data from the etomoxir experiments indicate that 200 μM treated cells have a truncated TCA cycle due to complex I inhibition . Increases in glycolysis and glutaminolysis are indicative of a change in mitochondrial activity [33] . Thus , we next examined oxygen consumption in whole cells after CPT1A knockdown . Unlike cells treated with 200 μM etomoxir , CPT1AKD cells had similar responses to respiratory inhibitors as wild-type cells with scrambled siRNA ( Fig 4B ) . Compared to control cells , however , CPT1AKD cells had a 40% increase in proton leak and a 60% decrease in ATP production . Taken together , CPT1AKD cells had a 70% decrease in mitochondrial coupling efficiency , which compromised their ability to efficiently use respiratory substrates for ATP production . Possibly to compensate for this loss in energy , CPT1AKD cells show increased basal and maximal respiration ( Fig 4C ) . These data are consistent with the observed increase in glucose , glutamine , and glutamate uptake ( Fig 4A ) . We also note that 200 μM etomoxir similarly inhibited respiration in CPT1AKD cells , which is consistent with etomoxir having off-target effects on the respiratory chain independent of CPT1A protein ( S17 Fig ) . To further examine mitochondrial dysfunction in CPT1AKD cells , we applied fluorescence imaging and electron microscopy ( EM ) . We first stained mitochondria with MitoTracker red , a positively charged fluorescent probe that accumulates as a function of membrane potential . We observed a significant increase in fluorescence intensity from MitoTracker red in CPT1AKD cells relative to controls , suggesting an alteration in mitochondrial membrane potential ( Fig 5 ) . Since interpreting this change with respect to increased or decreased mitochondrial membrane potential is complicated by the quenching effects of MitoTracker red at the concentration used , we also compared CPT1AKD and control cells with JC-1 staining [34 , 35] . JC-1 accumulates in the mitochondrial matrix as a function of the mitochondrial membrane potential . In the cytosol , JC-1 exists in its monomer form and fluoresces green . Upon its accumulation in the mitochondria , JC-1 forms aggregates that fluoresce red . Accordingly , depolarized mitochondria are characterized by a decrease in the red/green fluorescence intensity ratio [36] . In CPT1AKD cells , we found a decreased ratio of red J-aggregates to green J-monomers relative to control cells ( S18 Fig ) . As expected on the basis of our respiration measurements , these data are consistent with a depolarized mitochondrial membrane due to uncoupling in the CPT1AKD cells . Interestingly , upon CPT1A knockdown , we also observed multinucleated cells , which is a signature of cell-cycle arrest [37] . With electron microscopy ( EM ) imaging , we determined that more than 50% of the mitochondria in CPT1AKD cells had abnormal vesicular morphology compared to the well-defined cristae structure of control cells . Indeed , vesicular cristae shape has been associated with respiratory complex assembly and respiratory efficiency [38–40] . We did not observe abnormal mitochondrial morphology in etomoxir-treated cells ( S19 Fig ) , possibly due to a less complete inactivation of CPT1 compared to knockdowns . We note that although FAO is mostly inhibited in both BT549 cells treated with 200 μM etomoxir ( Fig 1B ) and in CPT1AKD cells ( Fig 3A ) , the isotopologue distribution patterns of citrate after U-13C palmitate labeling cannot be used to compare the level of CPT1A inhibition . This is because 200 μM etomoxir has the off-target effect of inhibiting complex I , which impairs the regeneration of NAD+ and thereby influences the oxidative degradation of U-13C palmitate . Pyruvate and uridine enable some cells lacking a functional mitochondrial electron transport chain to proliferate [41 , 42] . Thus , we sought to test whether pyruvate and uridine could rescue growth in CPT1A knockdowns with dysfunctional mitochondria . When BT549 cells with knocked down CPT1A were given pyruvate and uridine , their proliferation rate remained significantly less than that of controls ( S20 Fig ) . These results are consistent with CPT1A knockdown cells having a functional electron transport chain that can regenerate oxidized cofactors and suggest that their dysfunctional mitochondria impair cell growth by a different mechanism . Our data suggest that knocking down CPT1A affects cell proliferation through a mechanism that is independent of FAO . As one such potential mechanism , we considered the possibility that CPT1A plays an important structural function essential to the integrity of the mitochondrial membrane . To assess this hypothesis , we expressed CPT1A having G709E and G710E mutations in BT549 cells . The replacement of glycine residues 709 and 710 , which are part of the catalytic site , with glutamate abolishes CPT1A activity ( S21A Fig ) [43 , 44] . We refer to this catalytically dead CPT1A as CPT1Amutant . We also note that CPT1Amutant was resistant to any siRNA added to knock down wild-type CPT1A . This allowed us to knock down wild-type CPT1A in BT549 cells , without affecting CPT1Amutant expression . We found that expression of CPT1Amutant protein did not rescue cells in which wild-type CPT1A had been knocked down . Specifically , expression of CPT1Amutant did not restore proliferation or mitochondrial membrane potential in wild-type CPT1A knockdowns ( S21B–S21D Fig ) . These data do not support a structural role for CPT1A that is independent of FAO . As another mechanism for how CPT1A may influence cell proliferation independent of FAO , we considered the possibility that CPT1A mediates transport of long-chain fatty acids into the mitochondria for anabolic purposes . That is , instead of oxidizing long-chain fatty acids transported into the mitochondria by CPT1A for energy , we hypothesized that the carnitine shuttle provides an indispensable source of fatty acids to synthesize complex lipids during cellular proliferation [45] . To test our hypothesis , we isolated mitochondria from CPT1AKD and wild-type cells and applied lipidomic profiling to quantitate differences in mitochondrial lipids . Consistent with our prediction , many complex lipid species had decreased levels in CPT1AKD cells relative to wild-type cells ( Fig 6A ) . In our untargeted profiling experiment , 87% of the dysregulated lipids were decreased ( see S1 Table ) . We then quantified the change in concentrations of these altered lipid features , which included complex structural lipids such as phospholipids , sphingolipids , and cardiolipins ( Fig 6B ) . We also observed a nearly 2-fold decrease in complex signaling lipids such as lactosylceramide and glucosyl/galactosylceramides . Smaller decreases were found in other signaling lipids such as lysophospholipids and diacylglycerols . Whether they are a direct consequence of limited long-chain fatty acid availability or a downstream consequence of altered mitochondrial metabolism , these data suggest that CPT1A plays a role in regulating the levels of mitochondrial lipids .
In recent years , multiple cancers have been found to have increased expression of CPT1 and/or sensitivity to CPT1 inhibition [6 , 9] . In the conventional textbook picture of mammalian metabolism , CPT1 commits long-chain fatty acids to catabolic oxidation [46] . Thus , increased expression of CPT1 and/or sensitivity to CPT1 inhibition has been assumed to represent a demand for FAO and the ATP or NADPH provided . Our work here reveals 2 complications with this interpretation: ( 1 ) pharmacological inhibition of CPT1 with high concentrations of etomoxir , as is often used in cancer studies , leads to off-target effects , and ( 2 ) CPT1 influences the proliferation of several cancer cell lines independent of FAO . Treatment of BT549 breast cancer cells as well as several other cancer cell lines with 200 μM etomoxir significantly slowed cell proliferation , which is consistent with previous studies [18] . However , decreased cell proliferation at 200 μM etomoxir is not a result of inhibiting the primary target of etomoxir ( i . e . , CPT1 ) . Rather , 200 μM etomoxir inhibits complex I of the electron transport chain ( an off-target effect ) and leads to decreased cell proliferation independent of FAO . We note that 10 μM etomoxir efficiently blocked 90% of FAO and did not exhibit off-target effects on respiration; however , 10 μM etomoxir did not reduce BT549 cell proliferation . When most of FAO was inhibited with 10 μM etomoxir , BT549 cells adjust their uptake and utilization of other nutrients to compensate for the loss of FAO . These data indicate that FAO provides a dispensable source of ATP and reducing equivalents under standard cell-culture conditions . FAO generates acetyl-CoA , FADH2 , NADH , ATP , and potentially cytosolic NADPH . Importantly , all of these products can be derived from other nutrient sources without using CPT1 . Glucose , for example , can provide cytosolic NADPH via the pentose phosphate pathway and acetyl-CoA from glycolysis and the pyruvate dehydrogenase complex . FADH2 , NADH , and ATP can be obtained from the oxidation of glucose carbon through the TCA cycle . Similarly , reducing equivalents and ATP can be readily derived from glutamine [47] . Thus , while the products of FAO are highly valuable to a cell and may serve as a major energy source , they are not unique to the FAO pathway . Our results suggest that some cells , such as BT549 , can therefore compensate for the loss of FAO by adjusting nutrient uptake and utilization . Inhibiting approximately 90% of FAO by pharmacological inhibition of CPT1 did not affect the proliferation rate of BT549 cells , but genetic knockdown of CPT1A did . Moreover , genetic knockdown of CPT1A altered mitochondrial morphology and caused mitochondrial uncoupling , while pharmacological inhibition of CPT1 did not . These data together with the observations that acetate and octanoic acid did not rescue CPT1A knockdowns indicate that CPT1A has a function affecting cell proliferation that is independent of its role in FAO . We first considered a structural function of CPT1A as a scaffolding protein . However , expression of a catalytically dead CPT1A in BT549 cells in which wild-type CPT1A had been knocked down did not restore mitochondrial membrane potential . As another possible function of CPT1 that is independent of FAO , we considered the need to use CPT1 for purposes other than catabolic oxidation of lipids . Without CPT1 , cells cannot transport long-chain fatty acids into mitochondria , and therefore , downstream mitochondrial pathways using these substrates are impaired ( Fig 7 ) . Sources of long-chain fatty acids ( or long-chain fatty acyl-CoAs ) inside the mitochondria that do not rely on the CPT1 transport system are limited [48–50] . Complex lipids synthesized in the endoplasmic reticulum can be transported to the mitochondria and deacylated to make long-chain fatty acids [51 , 52] , or long-chain fatty acids can be generated in the mitochondrial matrix by type II mitochondrial fatty acid synthesis , a pathway that resembles fatty acid synthesis in bacteria [53] . Although the fates of long-chain fatty acids generated by these processes remain poorly understood , disrupting mitochondrial fatty acid synthesis slows cell growth , influences mitochondrial phospholipid composition , and alters mitochondrial morphology [54–58] , phenotypes which are highly consistent with those that we observed here with CPT1A knockdown . One possible explanation for these findings is that long-chain fatty acids generated in mitochondria are involved in phospholipid side-chain remodeling [54] . The de novo synthesis of cardiolipin in the mitochondria , for example , is followed by cycles of deacylation and reacylation . This remodeling process is essential to mitochondrial structure and function and , at least in part , uses acyl-CoA substrates in the mitochondrial matrix [59 , 60] . Another possible demand for long-chain fatty acids in the mitochondria is protein acylation , which may be used for protein anchoring , cell signaling , or protein trafficking . Although acylation of mitochondrial proteins remains largely unexplored , many mitochondrial proteins have been shown to be modified with long acyl chains in the mitochondrial matrix [61 , 62] . It is important to note that any anabolic demand for long-chain fatty acids transported by CPT1A in BT549 cells is likely to be low , since pharmacologically inhibiting most of CPT1 activity with low concentrations of etomoxir does not result in decreased cell proliferation or mitochondrial dysfunction . Interestingly , the demand for mitochondrial fatty acid synthesis is also low , but its disruption similarly results in decreased cell proliferation and mitochondrial dysfunction [54] . Our results therefore suggest that , like mitochondrial fatty acid synthesis , the CPT1 system may provide an indispensable source of long-chain fatty acids in the mitochondria to support processes that do not demand much carbon ( such as phospholipid remodeling and protein acylation ) but are essential to healthy mitochondrial function and cancer cell proliferation . We also point out that the results obtained for the cancer cells studied here are unlikely to be generalizable to all cancer cells; however , they demonstrate that additional evidence independent of CPT1 is necessary to implicate FAO as an antitumor target .
|
Oxidation of long-chain fatty acids inside of the mitochondrial matrix provides an essential source of energy for some cells . Since long-chain fatty acids cannot freely pass into the mitochondrial matrix , they rely on a protein called carnitine palmitoyltransferase I ( CPT1 ) for transport . Prior research has found that many tumors exhibit increased expression of CPT1 and/or sensitivity to CPT1 inhibition by a drug called etomoxir . These findings have led to thinking that cancer cells rely on fatty acid oxidation for energy . Here we present data that indicate otherwise , showing that inactivation of fatty acid oxidation has no effect on the proliferation of at least some cancer cell lines . Instead , these cells alter their utilization of other nutrients ( such as glutamine ) to compensate for the loss of fatty acid oxidation . We describe 2 discoveries that provide new insight into the role of fatty acid oxidation in cancer and help rationalize previous results . First , etomoxir has the off-target effect of inhibiting complex I of the electron transport chain . Second , CPT1 has other cellular functions that are independent of fatty acid oxidation . We suggest that one such function may be importing long-chain fatty acids into the mitochondria for anabolic fates , rather than catabolic oxidation .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"chemical",
"compounds",
"gene",
"regulation",
"metabolic",
"processes",
"cell",
"processes",
"carbohydrates",
"organic",
"compounds",
"glucose",
"acidic",
"amino",
"acids",
"amino",
"acids",
"mitochondria",
"bioenergetics",
"cellular",
"structures",
"and",
"organelles",
"small",
"interfering",
"rnas",
"lipids",
"cell",
"proliferation",
"proteins",
"gene",
"expression",
"chemistry",
"short",
"reports",
"oxidation",
"metabolism",
"biochemistry",
"rna",
"cell",
"biology",
"nucleic",
"acids",
"organic",
"chemistry",
"genetics",
"glutamine",
"biology",
"and",
"life",
"sciences",
"chemical",
"reactions",
"fatty",
"acids",
"energy-producing",
"organelles",
"non-coding",
"rna",
"physical",
"sciences",
"monosaccharides",
"citric",
"acid",
"cycle"
] |
2018
|
Identifying off-target effects of etomoxir reveals that carnitine palmitoyltransferase I is essential for cancer cell proliferation independent of β-oxidation
|
Estimating the reduction in levels of infection during implementation of soil-transmitted helminth ( STH ) control programmes is important to measure their performance and to plan interventions . Markov modelling techniques have been used with some success to predict changes in STH prevalence following treatment in Viet Nam . The model is stationary and to date , the prediction has been obtained by calculating the transition probabilities between the different classes of intensity following the first year of drug distribution and assuming that these remain constant in subsequent years . However , to run this model longitudinal parasitological data ( including intensity of infection ) are required for two consecutive years from at least 200 individuals . Since this amount of data is not often available from STH control programmes , the possible application of the model in control programme is limited . The present study aimed to address this issue by adapting the existing Markov model to allow its application when a more limited amount of data is available and to test the predictive capacities of these simplified models . We analysed data from field studies conducted with different combination of three parameters: ( i ) the frequency of drug administration; ( ii ) the drug distributed; and ( iii ) the target treatment population ( entire population or school-aged children only ) . This analysis allowed us to define 10 sets of standard transition probabilities to be used to predict prevalence changes when only baseline data are available ( simplified model 1 ) . We also formulated three equations ( one for each STH parasite ) to calculate the predicted prevalence of the different classes of intensity from the total prevalence . These equations allowed us to design a simplified model ( SM2 ) to obtain predictions when the classes of intensity at baseline were not known . To evaluate the performance of the simplified models , we collected data from the scientific literature on changes in STH prevalence during the implementation of 26 control programmes in 16 countries . Using the baseline data observed , we applied the simplified models and predicted the onward prevalence of STH infection at each time-point for which programme data were available . We then compared the output from the model with the observed data from the programme . The comparison between the model-predicted prevalence and the observed values demonstrated a good accuracy of the predictions . In 77% of cases the original model predicted a prevalence within five absolute percentage points from the observed figure , for the simplified model one in 69% of cases and for the simplified model two in 60% of cases . We consider that the STH Markov model described here could be an important tool for programme managers to monitor the progress of their control programmes and to select the appropriate intervention . We also developed , and made freely available online , a software tool to enable the use of the STH Markov model by personnel with limited knowledge of mathematical models .
Soil-transmitted helminths ( STHs ) is a group of four species of intestinal worms: Ascaris lumbricoides , Trichuris trichiura , Necator americanus and Ancylostoma duodenale . The latter two species are indistinguishable by microscopic examination of the eggs and are therefore frequently reported collectively as “hookworms” . STH infections cause morbidity by adversely affecting nutritional status and impairing cognitive processes [1] . The World Health Organization ( WHO ) recommends preventive chemotherapy ( that is , the periodic administration of anthelminthic medicines ) as a public health intervention for the control of STH [2] . The aim of the intervention is to reduce the number of individuals harbouring infections of moderate and heavy intensity and therefore to control the morbidity caused by STH infection [3] . We previously evaluated the performance of a Markov model in predicting the changes in STH prevalence during the implementation of preventive chemotherapy in Viet Nam over 54 months . The model is described fully elsewhere [4] . Predicting the short- to medium-term ( 1–5 years ) changes in STH prevalence could facilitate the work of control programme managers in selecting the most appropriate intervention and in monitoring the ongoing outcomes of the programme . In all the datasets used in the present study ( i . e . those from the studies used to calculate the model parameters and those from control programmes used to evaluate the prediction capacity of the model ) , Kato-Katz was the only method used to collect parasitological data . We understand the poor precision of the diagnostic methods traditionally used to monitor STH control programmes ( i . e . Kato-Katz ) especially when intensities of infection are low [1] . We recognize that the prevalence would have been higher if measured with a more precise method than Kato-Katz , but we wanted to provide programme managers with a tool capable of interpreting the imperfect data collected during routine control activities . The model requires pre-control data on the distribution of intensity of infection in the population , defined as the proportion of the population with zero eggs , light intensity , moderate intensity and high intensity of infection as defined by WHO [2] for each of the three STH parasites ( Table 1 ) . The Markov nomenclature refers to these classes of intensity as “condition states” ( CS ) . Initially , each individual in a specified population is categorized into one of the four CS but may transition to another CS after the first year of intervention: the changes between these four CS are defined as transition probabilities ( TP ) ; the 16 TP corresponding to each of the possible changes are captured in a four-by-four TP matrix . Fig 1 illustrates the TP and the CS for hookworm infection in Uganda resulting from the distribution of albendazole once a year among 1423 school-aged children individually followed for 3 years [5] . A transition probability matrix set ( TPMS ) is composed of three TP matrices ( one matrix for each STH species ) . The Markov model assumes the “stationarity” of the TP , in the sense that if the intervention remains the same ( e . g . albendazole is administered once a year ) the TPMS will also remain constant throughout the duration of the control programme . For example , Fig 1 shows that if 37% of heavy-intensity hookworm infections become light-intensity infections after the first year of preventive chemotherapy , the same proportion ( 37% ) of the remaining heavy-intensity infections will become light-intensity infections after the second year of the intervention . This assumption is based on: Additionally , we assume that the pharmacological interventions conducted in only a part of the population ( i . e . school-aged children ) do not reduce significantly the faecal contamination of the soil in the short term . In a previous study the Markov model with constant TPMS predicted changes of prevalence in Viet Nam at four time-points ( 6 , 12 , 30 and 54 months ) for each STH infection with an average precision of 1 . 7 percentage points [4] . In this study we aimed to test the model’s capacity to predict changes in STH prevalence during implementation of control programmes in other countries . One of the limitations of the model in its initial form is the requirement for a detailed dataset to calculate the TPMS: the minimum requirement to run the model is individual data ( including the intensity of infection for each STH parasite ) of at least 200 individuals , collected before and one year after the intervention . However , the epidemiological data available in STH control programmes are often limited , which can thus limit the application of the model . We therefore adapted the Markov model to also allow predictions in cases when a more limited amount of data is available . For this purpose we developed two simplified models in addition to the model initially developed and now titled Original Model ( OM ) : Table 2 summarizes the characteristics of the three models . We intend the model to be a simple tool to be used by programme managers in developing countries , and we therefore designed a software tool to enable its use by personnel with limited familiarity with computer modelling .
The OM was developed and tested on data from Viet Nam with satisfactory results . The model calculates the TP from individual data at baseline and one year after the intervention . The mathematical aspects of this model are presented in S1 Additional File . The SM1 was developed to run when one-year follow up data are not available and the calculation of TP is therefore not possible from the programme data . We developed 10 “standard” TPMS , one for each possible preventive chemotherapy intervention ( Table 3 ) for use instead of TPMS from programme data . For four types of intervention the TPMS were directly measured from control programmes for which baseline date and one-year follow-up data were available: For the other five types of intervention ( TPMS 5 , 6 , 7 , 8 and 10 ) we could not measure the TP directly but we estimated the impact of the intervention applying two assumptions: Table 3 summarizes the characteristics of the standard TPMS and the field study data used to develop them . The mathematical aspects of this model are presented in S1 Additional File . The SM2 was developed to run when baseline data relating to the intensity of infection are not available and the calculation of the CS is therefore not possible from the programme data . The model also assumes that one-year follow up data are not available and uses the TPMS to run the SM1 . Equations predicting the relationship between baseline prevalence and CS at baseline were developed using data from 18 countries; details of the methodology are published elsewhere [14] . The mathematical aspects of this model are presented in S1 Additional File; S2 Additional File , presents the values of coefficients used for Simplified Model 1 ( SM1 ) and Simplified Model 2 ( SM2 ) . The two simplified models ( SM1 and SM2 ) do not calculate specific TPMS for the control programme but apply standard TPMS that are calculated from programmes using a similar intervention; for example , to estimate the change in STH prevalence in Myanmar , the TPMS developed from the data from Viet Nam ( TPMS1 ) were applied because the intervention in the two countries is the same ( albendazole two rounds per year ) ; to predict the change in prevalence in Cambodia the TPMS developed from the data from Zanzibar were applied ( TPMS4 ) . Programme coverage ( that is , the percentage of individuals treated in the target group ) is an important parameter to be considered . If , for example , only 50% of the school-aged population is treated , we would expect changes in the intensity of STH infection only in the individuals that have been treated . For this purpose our model offers the functionality to use the coverage of the intervention to modify the output ( the TP matrices would then be applied only to a fraction of the individuals corresponding to the coverage ) . During the development of the model we observed that when the entire population is treated ( as is the case when albendazole is distributed for control of lymphatic filariasis ) , the predicted prevalence was consistently higher than the observed values; this discrepancy between the two predictions was increasing at the increase of the baseline prevalence . We interpret this discrepancy as a progressive decrease of reinfection due to the reduction of soil contamination consequent to administration of anthelminthics to the entire population . We adjusted for this factor by including a standard modification , dependent on the initial level of prevalence: if baseline prevalence is > 80% the standard modification becomes active after 5 years of follow up; for lower initial prevalence the standard modification becomes active after 3–4 years of follow up . This modification is only applied when the entire population is treated . We hypothesize that the simplified models ( SM1 and SM2 ) , despite using standard TPMS , could predict with sufficient precision the changes in STH prevalence over 6 years . To test the performances of the models we collected data from STH control programmes reporting the prevalence collected after a number of years of preventive chemotherapy from published and unpublished reports to extract the STH baseline epidemiology , the drug or drug combinations used , the type of administration ( community or school-based ) and the frequency of the intervention . These programme data are from countries different than the ones used to define the model parameters . In total we obtained information from 45 follow-up surveys ( with their respective baselines ) in 14 countries [5 , 15–31]: for three countries control programme managers provided the data [5 , 15 , 31]; for the remainder the datasets were obtained from published scientific literature [16–30] . Some baseline data corresponded to more than one follow-up: on average each baseline corresponded to 1 . 7 follow ups ( range 1–5 ) . The total number of individuals from whom data were collected was 36 108 , with an average sample size of 508 for each survey ( range 50–2885 individuals ) . Table 4 presents details of the datasets used to test the model . The datasets used to develop the standard TPMS ( Table 3 ) or to estimate the CS [6] were not used for the evaluation of the model predictions . Programme coverage is an important parameter to consider when predicting changes in STH prevalence; however , coverage information was missing from the reports [16–30] and therefore when testing the models we assumed that the coverage of all the interventions exceeded 80% ( i . e . the standard assumption in the model ) . The time interval between collection of baseline data and follow-up data ranged between 6 months and 13 years ( mean 3 . 2 years ) : in 24 . 4% of control programmes the interval between baseline and follow up was one year; in 37 . 5% the interval was between 2 and 4 years; in 35 . 6% the interval was 5–6 years; and in 2 . 3% the interval was more than 6 years . At baseline the average prevalence of A . lumbricoides infection was 34% ( range , 0–99% ) , of T . trichiura infection 27% ( range , 0–100% ) and of hookworm infection 45% ( range , 0–99% ) . At follow-up the observed average prevalence of A . lumbricoides infection was 13 . 9% ( range , 0–75% ) , of T . trichiura infection 12 . 8% ( range , 0–65% ) and of hookworm infection 14% ( range , 0–58% ) . Some 53 . 3% of control programmes used albendazole , 24 . 3% used mebendazole and the remainder used a combination of drugs which were being used for the control of lymphatic filariasis ( albendazole and either ivermectin or diethylcarbamazine ) . We tested the different forms of the model ( OM , SM1 and SM2 ) according to the kind of the data available from each programme: We developed two indicators to evaluate the accuracy of the model predictions for each dataset: To our knowledge , no other models predicting STH prevalence changes are available , so we assumed that a model predicting more than 70% of the cases within 10 percentage points would yield satisfactory results . The study was conducted on secondary data , and all data were recorded and analysed anonymously .
Details of the 10 standard TPMS are listed in S3 Additional File; each TPMS contains three TP matrices ( one matrix for each STH species ) . The nine multinomial curves and regression coefficients used to estimate the CS when the parasite prevalence is known ( but not the classes of intensity ) are published elsewhere [6] . The curves and coefficients allowed us to display graphically and estimate mathematically the prevalence of heavy , moderate and light intensity infections for each STH parasite , and therefore to run the model in cases where baseline information on CS was not available . In general the mean discrepancy for any of the three STH parasites between the observed and the predicted prevalence was 3 . 16 percentage points ( SD , 4 . 162 ) for OM , 4 . 8 percentage points ( SD , 6 . 075 ) for SM1 and 8 . 4 percentage points ( SD , 12 . 484 ) for SM2; these differences are marginally statistically different ( p = 0 . 0301 ) . Table 5 summarizes the mean discrepancy between predicted and observed prevalence values for each parasite and for each of the models tested . The accuracy of the models ( proportion of predictions within 10 percentage points ) for any of the STH parasites was 94% for OM , 84% for SM1 and 72% for SM2 . In general the models preformed best for T . trichiura ( 43% of predictions within 1 percentage point ) and least well for hookworm ( only 26% of predictions within 1 percentage point ) . Table 5 presents details of the accuracy of the predictions by parasite and by model and Fig 2 illustrates for each dataset the observed prevalence and the prevalence predicted by OM , SM1 and SM2 . We also compared the precision of the predictions of OM , SM1 and SM2 only on the datasets from Viet Nam and Uganda , since these were the only datasets that were tested on all three models . Furthermore , this analysis ( data not presented ) showed a better performance of OM than SM1 and SM2 for all the three STH parasites; the differences are , however , not statistically significant . We also evaluated the performance of the model using the Root Mean Square Error ( RMSE ) [32] . For OM the RMSE value was 0 . 04 ( range of prediction discrepancy , 0–0 . 14 ) , for SM1 the RMSE value was 0 . 11 ( range of prediction discrepancy , 0–0 . 27 ) and for SM2 the RMSE value was 0 . 15 ( range of prediction discrepancy 0–0 . 60 ) , showing a precision progressively declining from OM1 to SM1 and SM2 . In the datasets where the intensity of infection was reported over time [5 , 10] , the model could predict with less than 5% discrepancy the prevalence of the different classes of intensity . The software was designed to facilitate the application of the principle of the Markov model by managers of STH control programmes and is available online at: https://github . com/namkyodai/STHpredictor The software features are described in the S4 Additional File .
The tool was developed mainly for use in STH control programmes by managers wishing to: It is important that programme managers understand the limitations of these models . Here , we do not aim to model explicitly the many dynamic elements that will affect the level of transmission , such as the rates of contamination of the environment , water contact , parasite acquisition , and any age-related and immune-related processes that may affect these rates . Instead , the modelling approach is to implicitly combine all of these elements into a composite measure of transmission . Using the current approach , it is not possible to decipher which of the contributing factors leads to any lower than expected reductions in infection . Rather , where infection levels are observed to not being reduced as expected , this should prompt further investigation as to the underlying cause . The model’s predictive capacity was lowered at the extreme upper end of the observed prevalence range . As mentioned above , the SM2 model appeared to predict better in case prevalence is less than 90% . In case of very high prevalence such as that observed in Sri Lanka in our dataset ( > 98% prevalence ) , where it is likely that the proportion of individuals with high intensity of infection was underestimated for the CS . These are relatively rare values of infection in large-scale control programmes , and therefore fewer such data points are available to test the models . All versions of the model predicted the expected prevalence of hookworms less precisely than for the other STH species . This is probably because hookworms are in reality two different species with different sensitivity to anthelminthics but were constantly reported collectively as “hookworms” in all the studies we considered . For the purposes of medium- to short-term estimates ( 1–5 follow up years after baseline ) we consider that models provide useful information for control programme managers . Further testing of the model against additional datasets would allow a better assessment of the predictive capacity of the model . We therefore invite researchers and managers of STH control programmes to download the β version of the software , to test it directly with their data and to report back to the authors . We are now refining the software by linking it to a mapping tool to enable the development of maps illustrating the expected epidemiological changes as well as to estimate the costs and anthelminthics needed using population data and unit costs . Initial prototypes of this tool are being tested .
|
Several million children are periodically dewormed to prevent the impairments to health and the economy caused by soil-transmitted helminths in endemic communities . It is important that managers of STH control programmes be able to anticipate the impact of the control measures on the prevalence of the diseases for two main purposes: to select the most appropriate intervention; and to be alerted in case of changes in prevalence that are less favourable than the expected values . We therefore developed a model that predicts such changes , and also included modifications to enable its use in cases where a limited amount of data is available ( SM1 and SM2 ) . We retrospectively tested the performances of the model on 26 control programmes comparing the prevalence predicted by the model with those observed during programme implementation: in 77% of cases the prevalence predicted by the original model was within five percentage points of the observed values ( 69% in SM1 and 60% in SM2 ) . We consider the performances of these models to be satisfactory . We designed a free online software to facilitate the use of these models by programme managers who may not be confident with modelling procedures .
|
[
"Abstract",
"Introduction",
"Material",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"markov",
"models",
"tropical",
"diseases",
"geographical",
"locations",
"parasitic",
"diseases",
"animals",
"data",
"management",
"ascaris",
"ascaris",
"lumbricoides",
"mathematics",
"tanzania",
"neglected",
"tropical",
"diseases",
"africa",
"computer",
"and",
"information",
"sciences",
"vietnam",
"probability",
"theory",
"people",
"and",
"places",
"helminth",
"infections",
"asia",
"nematoda",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"soil-transmitted",
"helminthiases",
"organisms"
] |
2016
|
Markov Model Predicts Changes in STH Prevalence during Control Activities Even with a Reduced Amount of Baseline Information
|
Alkhurma hemorrhagic fever virus ( AHFV ) and Kyasanur forest disease virus ( KFDV ) cause significant human disease and mortality in Saudi Arabia and India , respectively . Despite their distinct geographic ranges , AHFV and KFDV share a remarkably high sequence identity . Given its emergence decades after KFDV , AHFV has since been considered a variant of KFDV and thought to have arisen from an introduction of KFDV to Saudi Arabia from India . To gain a better understanding of the evolutionary history of AHFV and KFDV , we analyzed the full length genomes of 16 AHFV and 3 KFDV isolates . Viral genomes were sequenced and compared to two AHFV sequences available in GenBank . Sequence analyses revealed higher genetic diversity within AHFVs isolated from ticks than human AHFV isolates . A Bayesian coalescent phylogenetic analysis demonstrated an ancient divergence of AHFV and KFDV of approximately 700 years ago . The high sequence diversity within tick populations and the presence of competent tick vectors in the surrounding regions , coupled with the recent identification of AHFV in Egypt , indicate possible viral range expansion or a larger geographic range than previously thought . The divergence of AHFV from KFDV nearly 700 years ago suggests other AHFV/KFDV-like viruses might exist in the regions between Saudi Arabia and India . Given the human morbidity and mortality associated with these viruses , these results emphasize the importance of more focused study of these significant public health threats .
Alkhurma hemorrhagic fever virus ( AHFV ) is a variant of Kyasanur Forest disease virus ( KFDV ) , and like KFDV , is a member of the mammalian tick-borne encephalitis group [family Flaviviridae , genus Flavivirus] . AHFV and KFDV cause similar disease syndromes in humans marked by fever , severe headache , myalgia and arthralgia , followed in a subset of cases by a hemorrhagic syndrome or diffuse neurological sequelae [1]–[4] . Recent epidemiological studies of AHFV outbreaks describe case fatality ratios close to those associated with KFDV infection ( 2–10% ) , suggesting that previous estimates of 25% might have been exaggerated due to unrecognized asymptomatic or mild cases [5] . Beyond supportive care , there is no specific treatment for either AHFV or KFDV infections . AHFV and KFDV have distinct geographic ranges in Saudi Arabia and India , respectively ( Figure 1 ) . AHFV was isolated in 1994 in Makkah , Saudi Arabia , from the blood of a fatally infected butcher [6] . Since then , AHFV cases have been confirmed in Jeddah , Jizan , and Najran , and most recently , outside of Saudi Arabia near the Egypt-Sudan border in 2010 ( Figure 1 ) [7] . Some early reports used an alternate spelling of Alkhumra [1] , [2] . Human AHFV infections have been associated with tick bites , and AHFV has been isolated from an Ornithodoros tick in Jeddah [8] , and Ornithodoros savignyi and Hyalomma dromedarii ticks in Najran [9] . However , another common risk factor for AHFV infection appears to be close contact with domestic animals , particularly sheep and camels [5] , although no disease has been reported in livestock or other animals . The host range of KFDV is quite different from AHFV; KFDV was first identified as the cause of nonhuman primate die-offs ( Presbytis entellus and Macaca radiata ) as well as a concurrent outbreak of fatal human disease in the Kyasanur forest region of India in 1957 . All known confirmed laboratory cases of KFD in people have been associated with contact with ticks ( primarily Haemaphysalis spp . ) , activities within the Shimoga forest region , or through laboratory infections . Despite apparent differences in their hosts and geographic ranges , AHFV and KFDV share high sequence identity [10] , [11] . Their positive-sense RNA genomes are approximately 11 kb in length and encode a single 3416 amino acid polyprotein that is post-translationally cleaved into a total of 3 structural ( C , M and E ) and 7 nonstructural ( NS1 , NS2a , NS2b , NS3 , NS4a , NS4b , and NS5 ) proteins . Given the notable genetic similarity and the later emergence of AHFV in Saudi Arabia , it has been speculated that AHFV arose following an introduction of KFDV from India . Previous phylogenetic analyses of AHFV [12] and KFDV [10] have relied on partial gene sequences , including regions of the structural envelope ( E ) , the RNA-dependent RNA polymerase ( NS5 ) , and the viral protease/helicase ( NS3 ) . These studies indicated a recent divergence of AHFV and KFDV , occurring sometime between 1828 and 1942 , presumably spawned by the introduction of KFDV into Saudi Arabia . However , given that the vector ecology , mammalian hosts and ecological niche of AHFV differ markedly from KFDV , it is possible that a longer period of divergent evolution between AHFV and KFDV could explain these significant biological differences . A more complete understanding of the evolutionary history of these viruses would provide insight into the circumstances surrounding their emergence . This is particularly important , as AHFV and KFDV are serious public health threats , and the recent identification of AHFV in Egypt demonstrates the potential for either viral spread and range expansion , or a larger range than previously thought . To gain insight into the relationship between AHFV and KFDV , we sequenced the full-length genomes of 16 AHFV and 3 KFDV isolates and analyzed those with two existing sequences available in GenBank . Our analyses revealed a higher overall diversity amongst AHFV strains than previously thought , particularly within the tick population . Surprisingly , these analyses indicated a much earlier divergence of AHFV from KFDV approximately 700 years ago , suggesting AHFV and KFDV might have broader geographic ranges , and raises the possibility of closely related but undiscovered virus variants existing in the regions between Saudi Arabia and India .
Animal procedures in this study complied with institutional guidelines , the US Department of Agriculture Animal Welfare Act , and the National Institutes of Health guidelines for the humane use of laboratory animals . Procedures were approved by the Centers for Disease Control and Prevention ( CDC ) Institutional Animal Care and Use Committee ( IACUC ) . The virus isolates included in this study are listed in Table 1 . All AHFV samples were received from the Saudi Arabia Ministry of Health . Partial sequences of the AHFV tick pool isolates have been published earlier [9]; the human AHFV samples had not been previously reported . Viruses were isolated through inoculation of sucking mice . Animals showing clinical symptoms were euthanized and brain suspensions were prepared in Hanks balanced salt solution and clarified by low-speed centrifugation , and the supernatant was used to infect Vero E6 cells . The KFDV isolates had been cataloged by the National Institute of Virology ( NIV ) in Pune , India after isolation in 1957 [10] . These lyophilized isolates were reconstituted in RNase-free water and used to infect Vero E6 cells . The cells were incubated at 37°C for 5 days , and then supernatants were harvested for RNA extraction . To extract the viral RNA , cellular supernatant was added at a 1∶10 ratio to Roche Tripure isolation reagent ( Roche Applied Science ) . The Tripure mixture was transferred to new tubes and surface decontaminated for removal from a biosafety level ( BSL ) -4 laboratory to a BSL-3 facility ( both laboratories are registered and approved by the National Select Agents program ) . 250 µL chloroform was added to the Tripure mixture , the tubes vortexed and centrifuged at 10 , 000 rpm for 15 min . The aqueous layer was added to an equal volume of 70% ethanol . The remaining procedure was carried out per manufacturer's instructions . RNA was eluted in 50 µL of RNase-free water . Three primer pairs were used to produce overlapping products of approximately 4 kb in length ( Table 2 ) using the Invitrogen Superscript III One-step RT-PCR with Platinum Taq DNA Polymerase kit ( Invitrogen ) . Two isolates , KFDV W-377 and KFDV G11338 , were amplified from the 5′ end with an alternate forward primer ( 6F ) . For each reaction , 5 µl RNA was added to 25 µl of the 2x Reaction mix , 10 µM of each primer and 1 µL of Superscript RT-PCR Taq mix in RNase-free water . The one-step RT-PCR protocol was as follows: reverse transcription at 52°C for 30 min , followed by denaturation at 94°C for 2 min and 40 cycles of 94°C for 15 s , 58°C for 30 s , 68°C for 4 min , and a final extension of 68°C for 5 min . To confirm the 5′ and 3′ ends of KFDV and AHFV , 3′ RACE was performed on the viral RNA ( vRNA ) and viral complementary RNA ( vcRNA ) of one KFDV isolate ( KFDV P9605 ) and one AHFV isolate ( AHFV 809006 ) . A polyA tail was added to both vRNA and vcRNA using the A-Plus Tailing kit ( Epicentre Biotechnologies ) in 50 µL reactions containing 35 µL RNA , 5 µL 10× A buffer , 5 µL 10mM ATP , 0 . 5 µL RNase-out , 1 µL polyA polymerase , and RNase-free water . Reactions were incubated for 15 min at 37°C , cleaned using Qiagen RNeasy columns ( Qiagen Inc . ) , and eluted into 50 µL of RNase free water . A 3′ RT-PCR on vRNA and a 3′ RT-PCR on vcRNA were run separately , each with the Invitrogen 3′RACE-AP primer and a gene-specific primer ( Table 2 ) . Superscript III RT-PCR with Platinum Taq ( Invitrogen ) was used in 50 µL reactions containing 5 µL RNA with polyA tail , 2 µL gene-specific primer ( 3′ or 5′ ) , 2 µL oligo dT primer , 25 µL 2x buffer , 1 µL enzyme mix , and 15 µl RNase-free water . Reactions were run at 50°C for 30 min and 94°C for 2 min , and then through 35 3-step cycles of 94°C for 30 s , 56°C for 30 s , and 68°C for 1min , with a final extension of 68°C for 5 min . Results from the genomic 3′ RACE were used to design the terminal 3′ primer ( alkG1R , Table 2 ) . PCR amplified products were sequenced using ABI Big-Dye 3 . 1 chemistry with an ABI 3730XL sequencer ( Applied Biosystems ) . All available AHFV and KFDV isolates in GenBank were aligned and a panel of 52 sequencing primers was designed to encompass the entire genome . The chromatogram data were assembled and analyzed as described previously [13] . Approximately 85-90 reads were obtained for each genome , resulting in an average six-fold redundancy at each base position . The 16 AHFV and 3 KFDV full-length sequences were compared to 2 available AHFV sequences in GenBank: the prototype AHFV strain 1176 ( GenBank Acc . #AF331718 ) and the partial coding sequence of AHFV strain JE-7 isolated from a tick ( GenBank Acc . #DQ154114 ) , for a total of 18 AHFV and 3 KFDV . All sequence alignments and pairwise comparisons were completed with MacVector with Assembler 11 . 1 . 2 ( MacVector , Inc . ) . Estimation of the gene-specific ratio of nonsynonymous to synonymous mutations were completed using DataMonkey webserver [14]–[16] . Using ModelTest [17] , the General Time Reversible model with a gamma distribution ( GTR+G ) was selected as the appropriate nucleotide substitution model . Marginal likelihood analysis indicated a relaxed lognormal clock [18] and a constant population size model was appropriate for the dataset . A Bayesian coalescent phylogenetic analysis was used to determine rates of evolution and time to most recent common ancestor ( TMRCA ) using BEAST 1 . 4 . 7 and Tracer 1 . 4 [19] . For the final analysis , 40 , 000 , 000 generations were run to ensure effective sample sizes ( ESS ) greater than 200 . TreeAnnotator and FigTree 2 . 0 were used to build and visualize trees [19] . After completion of these full length genome analyses , small partial sequence fragments of the E and NS5 from an AHFV recently isolated in Egypt were published ( GenBank Acc . #HM629507 , HM629508 ) [7] . To build an inclusive phylogeny containing this new data , the analyses described above were rerun separately to include these partial gene fragments .
Here we describe the first AHFV/KFDV phylogeny produced using full-length genomes . To sequence the available 16 AHFV and 3 KFDV isolates ( Table 1 ) , primer pairs were designed to amplify 3 overlapping segments of approximately 4 kb each . These RT-PCR products span the entire 10 . 7 kb genome excluding the terminal 26 nucleotides ( nt ) of the non-coding 5′ untranslated region ( UTR ) . The terminal 33 nt of the 5′ UTR and the terminal 39 nt of the 3′ UTR of KFDV-P9605 and AHFV-809006 were identical using 3′ RACE on the vRNA and vcRNA . The full genome lengths of AHFV and KFDV differed by only 1 nt; a conserved A in position 10415 of the noncoding 3′ UTR of AHFV was absent in KFDV , resulting in genome sizes of 10775 and 10774 , respectively . Analysis of full-length AHFV and KFDV genomes ( Table 3 ) confirmed a high level of sequence identity between the viruses , with an overall diversity of 8 . 4% ( nt ) and 3 . 0% ( aa ) . The region showing the most nucleotide diversity ( 11 . 6% ) within AHFV isolates was the 23 aa ( 69nt ) C-terminus of the NS4A protein , referred to as the 2K peptide [20] . However , the 2K amino acid sequence was completely conserved amongst AHFV isolates . The ratios of nonsynonymous to synonymous substitutions were determined for each AHFV were determined and showed no strong indication of positive selection . The mean rate of molecular evolution was estimated to be 9 . 2×10−5 sub/site/year ( 95% highest posterior density [HPD]: 1 . 6×10−5 sub/site/year [low] to 1 . 9×10−4 sub/site/year [high] ) . The phylogeny clearly showed AHFV and KFDV as 2 separate lineages ( Figure 2 ) . The TMRCA for all AHFV isolates was recent , 84 years before 2009 ( ∼1925 ) [HPD range: 29-165 years] . All KFDV isolates shared a recent TMRCA as well , 75 years before 2009 ( ∼1934 ) [HPD range: 56–104] . Despite their genetic similarity and individual recent TMRCAs , our analyses revealed that AHFV and KFDV diverged approximately 687 years before 2009 [HPD range: 121-1487 years] . This analysis encompassed 18 AHFVs isolated from either southern or western Saudi Arabia . Eight AHFVs were derived from ticks , including one published previously [8] . The remaining 10 were derived from human cases , including the prototypic AHFV 1176 , also previously published [11] . The resulting phylogeny showed all known AHFV isolates fall into 3 sublineages ( Figure 2 ) . Sublineage I consisted entirely of 5 tick pool isolates from the Al Balad Magan camel market in Najran . All were isolated from Ornithodoros savigyni ticks and had a pairwise diversity of 0 . 5% . Sublineage II included two tick pool isolates taken at the same time from a nearby camel market , also in Najran , in the Al Mishaaliyia district . Despite the close proximity of the camel markets , the genetic variation between tick pool isolates of Sublineage I and II was 1 . 1% , slightly higher than the overall diversity seen within all human AHFV isolates ( 0 . 7% ) . The tick pool isolates of Sublineage II came from different tick species ( one from Hyalomma dromedarii and one from O . savignyi ) but differed from one another by only 3 nt . There were no defining amino acid changes found between human and tick isolates . Sublineage II includes isolates from 1994-2009 , the full range of available full-length genomes , as well as the small gene fragments from the 2010 Egyptian isolate . Sublineage III encompasses isolates taken from ticks and humans in Makkah , Jeddah and Jizan . In summary , the genetic differences between AHFV isolates did not obviously correlate with geographic , temporal or host origin .
The emergence of AHFV in the mid-1990s and its remarkable genetic similarity to KFDV suggested that AHFV arose from a recent introduction of KFDV from India into Saudi Arabia . Using partial E and NS5 sequences of KFDV , the AHFV/KFDV divergence was estimated to have occurred 68 years ago [10] . Similarly , Charrel et al . , found a TMRCA of 182 years ago with partial E , NS3 , and NS5 sequences of AHFV [12] . Using only the concatenated E and NS5 sequences of our isolates ( as in [10] ) , we also found a recent time of divergence ( ∼100 years before 2009 ) . However , analyzing full-length genome sequences , we show 2 distinct lineages , with KFDV and AHFV diverging approximately 687 years ago . Analyzing full-length genomes allowed us to incorporate the full diversity of AHFV isolates , and therefore to provide a more accurate estimate of an older common TMRCA . The use of full genomes also revealed a slower rate of molecular evolution ( 9 . 2×10−5 substitutions/site/year ) than determined previously with partial sequences of KFDV [10] . This may be explained by the slightly higher NS5 and E diversity in KFDV relative to the rest of the genome ( Table 3 ) , however , the limited number of KFDV isolates available for this study might not be fully representative of the extant KFDV diversity . In accordance with the evolutionary rate patterns for vector-borne flaviviruses [21] , the AHFV/KFDV rate is slower than mosquito-borne flaviviruses , including Dengue virus ( DENV ) [22] and yellow fever virus ( YFV ) [23] and more similar to tick-borne flaviviruses , including the closely related tick-borne encephalitis virus ( TBEV Siberian subtype , 1 . 64×10−4 sub/site/year ) [24] . The evolutionary rate is also similar to that of an unrelated tick-borne member of the Bunyaviridae family , Crimean-Congo hemorrhagic fever virus ( 0 . 58–1 . 52×10−4 sub/site/year ) [25] . In our analyses , the sublineages of AHFV isolates did not show clustering by geography or date of isolation . Although these isolates primarily came from two distinct geographic areas , the Jeddah/Makkah region in the west and the Jizan/Najran region in the south , frequent travel of people and livestock between these regions effectively eliminates geographic boundaries . The short sequences available from the 2010 Egyptian isolate placed it within Sublineage II , the most inclusive of AHFV sublineages , possibly suggesting a relatively recent introduction to Egypt from Saudi Arabia . Alternatively , the limited sequence data available from the Egyptian isolate is not representative of extant diversity within East Africa and the full-length sequence could indicate a different evolutionary history . The narrow range of time encompassed by the viral isolates ( ∼15 years ) and the slow evolutionary rate of AHFV may also explain the lack of detectable temporal clustering . The centuries between the divergence of the ancestral AHFV and KFDV ( ∼year 1322 ) , and the individual TMRCAs of AHFV ( ∼1925 ) and KFDV ( ∼1933 ) isolates , coupled with a slow evolutionary rate , indicate the 2 viruses evolved separately and AHFV is not a result of recent KFDV introduction in Saudi Arabia [10] . This separate evolutionary history spanning several hundred years and the vast geographic distance involved raises the possibility of an existing spectrum of thus far unknown tick-borne encephalitic/hemorrhagic viruses in the regions between Saudi Arabia and India . In support of such a viral spectrum in that region , Gould et al . [26] cite the presence of close AHFV/KFDV relatives , Karshi virus ( KSIV ) and Farm Royal virus ( FRV ) , in Uzbekistan and Afghanistan , respectively . A similar dispersal corridor of TBEV has been described from Eastern Europe westwards [27] , although more recent studies did not support such a pattern [28] . Dispersal of the ancestral viruses of AHFV and KFDV may have been accomplished through the movement of animals , including camels presumably carrying ticks , along the Silk Road , which by the 1300s stretched from Europe to China . The presence of potentially competent tick vectors in these regions [29] also supports this possibility of AHFV/KFDV-like viruses existing between Saudi Arabia and India . Although the viruses are primarily associated with two tick genera , AHFV with Ornithodoros and KFDV with Haemaphysalis , both have been isolated from multiple genera . At least 16 tick species have been shown to transmit KFDV [3] . The experimental finding of transovarial and transtadial transmission of KFDV in O . cruzi [30] , coupled with the ubiquitous nature of Ornithodoros spp . in India , Saudi Arabia and the intermediate regions , suggests these ticks may have played an important role in the spread of AHFV/KFDV-like viruses . Although the transmission of KFDV has been primarily associated with H . spinigera , a tick species not documented in Saudi Arabia , these ticks are present between Saudi Arabia and India [29] . Given the diversity of potential vectors and their extensive ranges , it is feasible that AHFV and KFDV , as well as other similar but thus far undiscovered viruses , are circulating more broadly throughout Saudi Arabia , India and beyond . The genomic diversity of AHFV was most apparent within isolates from the tick population . Ticks sampled on the same day in one city ( Najran ) harbor a more genetically diverse AHFV population than isolates obtained from human cases over the course of 15 years throughout Saudi Arabia . Similarly , an analysis of the European subtype of tick-borne encephalitis virus ( TBEV ) isolated from ticks shows relatively high variation in a limited geographic area and temporal period , perhaps due to importation of TBEV on migratory birds [31] , [ 32] . The notable genetic heterogeneity of AHFV in Najran may be explained similarly , as camel and other livestock traders often travel large distances to Najran to market their animals . Like all soft ticks , Ornithodoros spp . feed for only short periods , spending most of their time in burrows or nests , allowing areas such as a camel market to become a potential focus of AHFV infection and high virus genomic diversity . Presumably , like KFDV , AHFV can persist for long periods within a tick , and given the nature of RNA viruses , variants may evolve over time within the invertebrate hosts . The lower diversity within human cases may reflect limited transmissibility or reduced pathogenicity of some of these variants in people . In our analysis , the use of full-length AHFV and KFDV genomes demonstrated a deeper evolutionary history than suggested by previous partial genome analyses . The divergence of AHFV and KFDV almost 700 years ago indicates a long period of divergent evolution , and suggests that a range of as-yet undiscovered tick-borne hemorrhagic/encephalitic viruses could exist between Saudi Arabia and India . The notably high AHFV diversity found within tick populations , coupled with the extensive geographic range of competent tick vectors , raises the possibility of broader AHFV and KFDV geographic ranges , and is supported by the recent discovery of AHFV in Egypt . As AHFV and KFDV are both associated with significant human morbidity and mortality , the potential spread of these viruses should be of serious concern and warrants further study of these significant pathogens .
|
Alkhurma hemorrhagic fever ( AHF ) and Kyasanur Forest disease ( KFD ) viruses both cause serious and sometimes fatal human disease in their respective ranges , Saudi Arabia and India . AHFV was first identified in the mid-1990s and due to its strong genetic similarity to KFDV it has since been considered the result of a recent introduction of KFDV into Saudi Arabia . To gain a better understanding of the evolutionary history of AHFV and KFDV , we sequenced the full-length genomes of 3 KFDV and 16 AHFV . Sequence analyses show a greater genetic diversity within AHFV than previously thought , particularly within the tick population . The phylogeny constructed with these 19 full-length sequences and two AHFV sequences from GenBank indicates AHFV diverged from KFDV almost 700 years ago . Given the presence of competent tick vectors in the regions between and surrounding Saudi Arabia and India and the recent identification of AHFV in Egypt , these results suggest a broader geographic range of AHFV and KFDV , and raise the possibility of other AHFV/KFDV–like viruses circulating in these regions .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"infectious",
"diseases",
"public",
"health",
"and",
"epidemiology",
"rna",
"viruses",
"virology",
"viral",
"classification",
"biology",
"microbiology",
"evolutionary",
"biology",
"public",
"health"
] |
2011
|
Ancient Ancestry of KFDV and AHFV Revealed by Complete Genome Analyses of Viruses Isolated from Ticks and Mammalian Hosts
|
Knowledge of the foci of Plasmodium species infections is critical for a country with an elimination agenda . Namibia is targeting malaria elimination by 2020 . To support decision making regarding targeted intervention , we examined for the first time , the foci of Plasmodium species infections and regional prevalence in northern Namibia , using nested and quantitative polymerase chain reaction ( PCR ) methods . We used cross-sectional multi-staged sampling to select 952 children below 9 years old from schools and clinics in seven districts in northern Namibia , to assess the presence of Plasmodium species . The median participant age was 6 years ( 25–75%ile 4–8 y ) . Participants had a median hemoglobin of 12 . 0 g/dL ( 25–75%ile 11 . 1–12 . 7 g/dL ) , although 21% of the cohort was anemic , with anemia being severer in the younger population ( p<0 . 002 ) . Most of children with Plasmodium infection were asymptomatic ( 63 . 4% ) , presenting a challenge for elimination . The respective parasite prevalence for Plasmodium falciparum ( Pf ) , Plasmodium vivax ( Pv ) and Plasmodium ovale curtisi ( Po ) were ( 4 . 41% , 0 . 84% and 0 . 31% ) ; with Kavango East and West ( 10 . 4% , 6 . 19% ) and Ohangwena ( 4 . 5% ) having the most prevalence . Pv was localized in Ohangwena , Omusati and Oshana , while Po was found in Kavango . All children with Pv/Pf coinfections in Ohangwena , had previously visited Angola , affirming that perennial migrations are risks for importation of Plasmodium species . The mean hemoglobin was lower in those with Plasmodium infection compared to those without ( 0 . 96 g/dL less , 95%CI 0 . 40–1 . 52 g/dL less , p = 0 . 0009 ) indicating that quasi-endemicity exists in the low transmission setting . We conclude that Pv and Po species are present in northern Namibia . Additionally , the higher number of asymptomatic infections present challenges to the efforts at elimination for the country . Careful planning , coordination with neighboring Angola and execution of targeted active intervention , will be required for a successful elimination agenda .
Progress in malaria elimination efforts have intensified across the Southern African Development Community ( SADC ) with the aim of elimination by 2030 for most countries [1] , [2] , [3] , [4] . This means a knowledge of current levels of transmission as part of the malariogenic potential is required . The burden of malaria is still a challenge with sporadic epidemic outbreaks within the region annually , due to continual movements within and between countries [2] , [5] , [6] . So , the major task for these countries is the increased detection of foci of asymptomatic and symptomatic Plasmodium species infections with sustained surveillance efforts , for targeted intervention and prevention of epidemic outbreaks [7] , [8] , [9] . We previously reported on the presence of P . vivax asymptomatic infections in Botswana which has impacted the malaria elimination agenda for better [10] . Asymptomatic infections are largely submicroscopic under low transmission settings , so they are not seen either by microscopy or rapid diagnostic test ( RDT ) and form significant reservoirs for reinfections and transmission of the parasite , making elimination a difficult task [11] , [12] . Namibia is a member of the elimination eight group ( E8 ) of SADC , which has targeted malaria elimination by 2020 as part of the vision 2030 agenda to mitigate poverty ( WHO , World Malaria Report , 2017 ) . It has a population of 2 . 2 million over an area of 0 . 83 km2 [8] , [2] . A greater proportion of the population ( 55% ) live in the north , which is traditionally , the main malarious regions in the country [2] . Namibia has achieved significant declines in symptomatic cases of malaria due to effective control methods [8] . However , the continual migration of individuals to and from Angola ( a malaria endemic country ) in the North , poses a major challenge of importation of parasites , which confounds the efforts at elimination as it creates changes and uncertainties in the disease burden with hidden asymptomatic cases [13] , [7] . To assess the risk associated with persistent migrations , continual surveillance to identify the foci and dynamics of asymptomatic and symptomatic Plasmodium species infections in the country is critical for targeted interventional approaches [14] . A previous modelling using the Bayesian Model Based Geostatics ( MBG ) approach predicted that the three districts most likely to experience rebound or epidemics of malaria in the country were: Ohangwena , Kavango and Caprivi that borders Angola and Zambia in the north [1] , [15] , [16] . Here we report for the first time on the foci of P . vivax ( Pv ) and P . ovale curtisi ( Po ) asymptomatic and symptomatic infections and current levels of P . falciparum ( Pf ) asymptomatic infections in an active survey in the country using PCR ( nested and qPCR ) a more sensitive molecular method .
The study sites were selected with the support of the National Malaria Control Program ( NMCP ) of the Ministry of Health and Social Services ( MOHSS ) and Ministry of Education , Namibia . The sites were: Kunene , Omusati , Oshikoto , Ohangwena , Kavango West , Kavango East and Zambezi ( Caprivi ) ( Fig 1 ) . The sites are color coded to reflect the malaria incidence Map for these sites in 2015 , which had not been updated at the time of sample collection . The respective incidence rates were , more than 5 cases per 1000 people in Kavango East and West , 1–4 . 99 cases per 1000 people in Zambezi , Ohangwena and Omusati , while Oshikoto and Kunene had less than 1 case per 1000 people . Within each district the Ministry of Education primary school registers and MOHSS Clinic/Health post registries were used in a multi-staged sampling process to randomly select the schools and clinics from which the enrolment was done . Details of the sampling procedure are as previously described [10] . Briefly , in the first stage , districts with known malaria transmission profile based on the recommendations of the NMCP were purposely selected . In the second stage , towns within the district with variable malaria incidence rates were also purposely selected from the NMCP recommendations as shown in Fig 1 . This was to ensure that areas of high , moderate , low and sporadic transmissions including those closer to the Angolan border were captured in the sampling process . In the third stage , schools and clinics within each town/village were purposely selected using a two-stage clustering approach based on the population density and adequate cross-sectional representation of the communities to avoid any bias . In the final stage , participants were assigned numbers and enrolled based on informed parental consent and consent of heads of schools and clinics . The total number of samples derived for each district/town was in direct proportion to the estimated population density . The study was approved by the MOHSS Ethical Committee , Namibia . All parents/guardians provided informed consent on behalf of all participants . Where needed , assent was also obtained from the child before sample collection . The study enrolled a total of 952 individuals under 9 years old . Subjects enrolled from schools amounted to 591 while 361 were from Clinics and Health posts . Sampling collection was done from September 2016-October 2017 using a multistage sampling strategy as described previously . Fever was defined as subjects with axillary temperature of ≥ 37 . 2°C at the time of sample collection , while asymptomatic subjects were those without fever ( axillary temperature <37 . 2°C ) and without a history of fever in the preceding 72 hours . Sample collections were timed to cover the onset and peak of the malaria transmission season . Prior to the blood sample being drawn , a short questionnaire was administered for travel history and previous malaria illness in the past year , while basic information on age and sex were documented . An aliquot of 1 . 5–2 . 5 ml venous blood was collected into EDTA tubes and centrifuged at 3000 rpm for 5 minutes to separate the buffy coat , plasma and red blood cells into separate tubes . These were then stored at -20°C and later transferred to -80°C till analyzed . Hemoglobin ( Hb ) was measured using Hemocue ( Ängelholm , Sweden ) . Data were entered in an Excel data sheet and STATA v11 . 2 ( StataCorp , College Station , TX , USA ) was used for analysis . Descriptive statistics and appropriate measures of central tendency were provided for relevant demographic covariates . To describe differences between study sub-populations ( eg . different regions of residence , presence/absence of Plasmodium infection ) , continuous covariates were compared using linear regression or the student t-test and categorical variables were compared using logistic regression , the Chi square test , or Fisher’s exact test . For logistic regression analyses , odds ratios ( ORs ) were provided . For all point estimates , 95% confidence intervals ( CIs ) were provided . Anemia was defined as Hb < 11 . 0 g/dL . Statistical significance for all comparisons was set at p<0 . 05 and adjustment was not done for multiple comparisons in this exploratory study . P-values smaller than 0 . 001 were reported as p<0 . 001 .
A total of 952 children from 7 districts ( Fig 1 . ) were assessed for Plasmodium species infection . The median participant age was 6 years ( 25–75%ile 4–8 years ) . Participants recruited in Kunene ( 1 . 60 years younger , 95%ile 0 . 91–2 . 29 years younger ) , Ohangwena ( 1 . 24 years younger , 95%CI 0 . 74–1 . 74 years younger ) , and Omusati ( 1 . 20 years younger , 95%ile 0 . 66–1 . 75 years younger ) were all significantly younger than those participants recruited in Kavango East ( p<0 . 001 for all these comparisons ) . Overall , 52 . 6% were female . Participants had a median hemoglobin of 12 . 0 g/dL ( 25–75%ile 11 . 1–12 . 7 g/dL ) ; 21% of the cohort was anemic . There were significant differences in mean participant hemoglobin levels between different regions; those from Kunene ( 0 . 72 g/dL less , 95%CI 0 . 26–1 . 19 g/dL less , p = 0 . 002 ) and Ohangwena ( 0 . 65 g/dL less , 95%CI 0 . 31–0 . 99 g/dL less , p<0 . 001 ) had lower levels than children from Kavango East . The majority of Plasmodium infections were Pf ( n = 41 ) , with 8 Pv infections , 3 Po infections and no Pm infections ( Fig 2 ) . Coinfections were common; half of the children with Pv infections were co-infected with Pf and all the children with Po infections had Pf co-infections . Most of children with Plasmodium infection ( 63 . 4% ) were afebrile . All three children with Pv/Pf coinfections and two with Pf infection in Ohangwena , had previously visited Angola . The mean hemoglobin was lower in those with Plasmodium infection as compared to those who did not ( 0 . 96 g/dL less , 95%CI 0 . 40–1 . 52 g/dL less , p = 0 . 0009 ) . There were no differences in the age or gender distributions for those that did and did not have Plasmodium infections . There were clear differences in the proportions of participants infected in the different regions; the highest prevalence was seen in Kavango East ( 10 . 4% ) and there was a statistically lower prevalence seen in the regions of Kunene ( 0% , p<0 . 001 ) , Ohangwena ( 4 . 50% , OR 0 . 40 95%CI 0 . 17–0 . 94 , p = 0 . 04 ) , Omusati ( 2 . 70% , OR 0 . 13 95% CI 0 . 030–0 . 58 , p = 0 . 007 ) , and Oshana ( 2 . 63% , OR 0 . 17 95%CI 0 . 039–0 . 75 , p = 0 . 02 ) . There was no statistical difference between the prevalence of mixed infections in the different regions . Kavango region interestingly , is the most endemic for malaria from the 2015 incidence Map . Pv was localized within Omusati , Ohangwena and Oshana regions , while P ovale was seen in Kavango . The precise number and type of Plasmodium species for each region is presented in Table 1 . The overall parasite prevalence was 4 . 83% with Pv and Po accounting for 0 . 84% and 0 . 31% respectively . The two species therefore account for 1 . 1% of the parasite prevalence .
The study has demonstrated for the first time that the Pv foci of infection in Northern Namibia encompasses three regions: Omusati , Ohangwena and Oshana , while Po focus is in Kavango . The study affirms Kavango as remaining a focus for Pf infections in northern Namibia , as shown in the incidence map in 2015 and epidemic outbreaks in 2016 [7] . The transmission profile of the parasite shows that Pv only infections were all asymptomatic and in a relatively younger population , whereas Pf infections were made up of relatively older children and a high number of asymptomatics , spread across all the regions . A similar pattern of Pv infection was observed in our previous study in Botswana [10] . Reports from Mali , Senegal , Indonesia , Papua New Guinea ( PNG ) , Brazil and others also indicate that asymptomatic infections are commonly observed [22] , [23] . Since the population with Pv infections were also anemic compared with those predominantly in Kavango East with Pf infection , the findings reveal some basic differences in the infection dynamics of Pv and Pf in low transmission settings , and in the population in northern Namibia . Pv strictly infects reticulocytes [24] , so under anemic conditions where increases in reticulocyte counts occur , Pv infection will be facilitated . In addition , the relapses associated with the hypnozoites stage of Pv infection [25] , will increase anti-blood stage immune acquisition more rapidly overtime [26] , [27] , so that eventually , this will lead to a predominantly asymptomatic population [28] , [29] . It is known that all stages of Pv infection can develop in the asymptomatic state [30] to sustain transmission . On the other hand , in Pf infections where hypnozoite stages are absent , if transmission is persistent , individuals acquire immunity with exposure , so the period of immune acquisition is longer and older people become the carriers of asymptomatic infections [22] . The Pf asymptomatic infections point to sustained infections over the years within the population , that has enabled older children to acquire enough immunity to harbor parasites [2] . In a recent MGB modelling method used to predict which regions in Namibia are most likely to experience rebound epidemics , Ohangwena , Kavango and Caprivi ( Zambezi ) were cited as the major focal areas [7] , which is in absolute congruence with the present report . It appears that there is a “quasi-stable” malaria infection scenario within a low transmission setting , where infections although sporadic may be persistent . One should also take cognizance of the fact the there is an inherent genetic variability of parasites for each population that adds to the heterogeneity and so require more tailored interventions , with regards to elimination [31] . The elimination agenda for Namibia has a major challenge of perennial importation of parasite across the border with Angola . This was seen in the present study , with visits to Angola contributing to the Pv infections . The challenge significantly complicates the elimination process [1] . There is a need for active systematic investigations and understanding of the epidemiology of asymptomatic malaria of all species in low transmission settings with an elimination agenda . This can initiate in hotspots and hotpops . Asymptomatic malaria sustains malaria transmission all season and so form a formidable component of transmission as they are not targeted for clearance [32] . This could be a major obstacle for elimination in a scenario where the asymptomatic fraction of the population grows rather than diminish with time , as a result of acquired immunity . Pv and Po hypnozoite forms add to the complexity of dealing with asymptomatic infections with their sequestration in the liver and or bone marrow [33] , [22] , [34] . So , in low transmission settings , the epidemiology of Plasmodium infections can be heterogeneous , requiring a more thorough active assessment in children towards malaria elimination . It is now no longer valid that Pv infections do not occur in Africa where Duffy antigen negativity is predominant . Several reports indicate that Pv infections occur in Duffy negative individuals in Africa [35] , [36] . Clearly , the agenda for malaria elimination should not only be focused on Pf infections but done in parallel with non-falciparum malaria . We conclude that Pv asymptomatic infections and Po are present in northern Namibia as are asymptomatic infections of Pf . These introduce new paradigms in the elimination agenda for Namibia , that requires careful planning and thought for blocking transmission and aggressive targeting of all populations and species affected .
|
Namibia is a member of the SADC elimination 8 ( E8 ) group with a target to eliminate malaria by 2020 . This target stems from years of aggressive interventional strategies that has led to significant reductions in morbidity and mortality . The focus of this strategy is mainly on Plasmodium falciparum as the primary parasite species . Foci of transmission is found in the northern border with Angola and Zambia , which also carries the highest population density . Recently as part of the elimination efforts to predict areas likely to have rebound epidemics , three regions Ohangwena , Kavango and Zambezi were identified . In order to affirm these findings and decision-making process for intervention , we assessed the parasite prevalence in 7 northern regional sites for four Plasmodium species . We identified Pv and Po curtisi parasites in Omusati , Ohangwena and Kavango , as well as a significant number of asymptomatic Pf and Pv infections , part of which may be due to importation from neighboring Angola . As Namibia is targeting elimination by 2020 , careful thought and planning will be required to reach the goal .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"parasite",
"groups",
"medicine",
"and",
"health",
"sciences",
"namibia",
"education",
"plasmodium",
"angola",
"sociology",
"tropical",
"diseases",
"geographical",
"locations",
"social",
"sciences",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"parasitology",
"apicomplexa",
"protozoans",
"molecular",
"biology",
"techniques",
"africa",
"research",
"and",
"analysis",
"methods",
"malarial",
"parasites",
"artificial",
"gene",
"amplification",
"and",
"extension",
"schools",
"molecular",
"biology",
"people",
"and",
"places",
"eukaryota",
"polymerase",
"chain",
"reaction",
"biology",
"and",
"life",
"sciences",
"malaria",
"organisms"
] |
2019
|
Molecular detection of P. vivax and P. ovale foci of infection in asymptomatic and symptomatic children in Northern Namibia
|
How chaperones interact with protein chains to assist in their folding is a central open question in biology . Obtaining atomistic insight is challenging in particular , given the transient nature of the chaperone-substrate complexes and the large system sizes . Recent single-molecule experiments have shown that the chaperone Trigger Factor ( TF ) not only binds unfolded protein chains , but can also guide protein chains to their native state by interacting with partially folded structures . Here , we used all-atom MD simulations to provide atomistic insights into how Trigger Factor achieves this chaperone function . Our results indicate a crucial role for the tips of the finger-like appendages of TF in the early interactions with both unfolded chains and partially folded structures . Unfolded chains are kinetically trapped when bound to TF , which suppresses the formation of transient , non-native end-to-end contacts . Mechanical flexibility allows TF to hold partially folded structures with two tips ( in a pinching configuration ) , and to stabilize them by wrapping around its appendages . This encapsulation mechanism is distinct from that of chaperones such as GroEL , and allows folded structures of diverse size and composition to be protected from aggregation and misfolding interactions . The results suggest that an ATP cycle is not required to enable both encapsulation and liberation .
While many proteins can successfully fold independently and spontaneously in vitro [1] , folding within the cell is facilitated by molecular chaperones [2 , 3] . Trigger factor ( TF ) is a general ATP-independent chaperone [4 , 5] found in bacteria ( e . g . , E . coli ) [6] and chloroplasts [7 , 8] . Bound to the ribosome exit tunnel [9] , TF interacts with the emerging nascent chains , and shields them from interactions with other cellular components [10] . TF can also remain associated with polypeptide chains when they are released into the cytosol , and interact with fully folded proteins before their assembly into larger protein complexes [11] . At the ribosome exit tunnel , TF adopts an extended conformation necessary for unfolded substrate interaction [10] . Its dragon-like structure ( see Fig 1A ) is composed of three domains: N-terminal ( residues 1–149 ) ; Polypropyl Isomerase or PPIase domain ( residues 150–245 ) , also referred to as “Head”; and C-terminal ( residues 246–432 ) , which forms the cradle of TF . The C-terminal comprises of a long helical linker ( residues 246–302 ) , named “HA1-linker , ” with two arm-like extensions—“Arm1” ( residues 303–359 ) and “Arm2” ( residues 360–415 ) . A “Linker” ( residues 112–149 ) connects the core of N-terminal ( residues 1–111 ) to the Head . In separate studies , Singhal , et al . [12] , and Thomas , et al . [13] , have used molecular dynamics simulations to reveal a surprising structural flexibility enabled by the hinge-like motions of linkers , which can drive a structural collapse of TF in isolation . It has been suggested that its flexibility and promiscuous surface allow TF to interact with substrates of diverse compositions and sizes [8 , 11 , 14] . How TF interacts with substrates away from the ribosome remains poorly understood . TF is known to bind unfolded and partially folded protein chains [4 , 15] , and slow down or delay the onset of their folding [16 , 17] . Using NMR analysis , Saio et al . showed that the sites for TF’s ( mostly ) hydrophobic interactions with unfolded PhoA are distributed over multiple domains on TF [14] . With optical tweezers experiments [3] on single Maltose Binding Protein ( MBP ) molecules , Mashaghi et al . have shown that that TF not only interacts with unfolded protein chains , but also binds and stabilizes folded structures of different sizes within partially folded proteins [18] . They found that TF binding partially folded structures protects against aggregation interactions , and , hence , promotes folding to the native state . Thus , TF not only acts before the onset of folding , but also during folding as the peptide chain folds step-by-step into a fully folded protein structure . However , it remains unclear how TF binds and stabilizes protein folds at the molecular level , which lies at the basis of its ability to guide the protein chain during folding . Resolving this issue is challenging , given the transient nature of the complexes between TF and partially folded structures . In this work , we use molecular dynamics ( MD ) simulations to characterize the interactions and the structural dynamics of the TF-MBP complex . We study the interactions with TF in three distinct stages of MBP’s folding process: fully unfolded protein chains , partially folded states , and native MBP . Fig 1B , 1C and 1D show the structural representations of native MBP and its partially folded states that were employed in this work . All-atom MD simulations of chaperone-assisted protein folding are computationally very expensive due to the large system size and the long time scales connected to relaxation . Using coarse-grained forcefields ( 1 bead per amino acid ) , O’Brien et al . [19] investigated protein folding at the ribosome , while extending the nascent chain . They showed that TF facilitates folding by decreasing the rate of structural rearrangements within the nascent chain , and increasing the effective exit tunnel length . While greatly extending time scales , this study missed the atomistic details of the interactions of TF with the substrate protein , as well as changes within the substrate proteins . The experimentally observed interactions between TF and partially folded states [18] provided an opportunity to study the mechanisms of interactions between TF and its substrate during the folding process . Therefore , we performed all atom MD simulations on MBP truncates that correspond to the experimentally observed partially folded states—namely , P1 ( residues 7–62 ) and P2 ( residues 112–248 ) [18 , 20] ( see Fig 1C and 1D ) . These partially folded truncates owing to their transient stability , provide good model substrates at lower computational cost . In this study , we focus primarily on the dynamics of the smallest partial fold—P1 . In addition , we employ steered molecular dynamics simulations on folded P1 to understand the stabilization effect of TF by comparing the potentials of mean force of P1’s unfolding in presence and absence of TF . We found that TF interacts with folded structures with the ends of its flexible appendages , especially N-terminal , thus forming a “Touching complex” . TF can also leverage its flexibility to wrap its appendages around the folded structures ( P1 ) , forming the more stable “Hugging complexes , ” while other MBP segments remain unfolded . This binding is the strongest with unfolded protein chains , and weakens gradually with the progressive folding of the polypeptide . It weakens significantly upon the formation of fully folded MBP , which likely plays a role in the post-folding release of the protein . Our steered MD simulations showed that breaking the end-to-end β–sheet hydrogen bonds of P1 in these complexes requires more work than when P1 is in isolation . This suggests that by wrapping around partially folded structures and establishing multiple favorable contacts , TF stabilizes these partial folds . In independent MD studies , Singhal , et al . [12] , and Thomas , et al . [13] found that the Arm1 plays an important role in TF’s collapse by strongly interacting with the PPIase domain . Consequently , Thomas , et al . speculated that the TF’s collapse may adversely affect its interactions with unfolded chains . Our results show that the binding of unfolded polypeptides is indeed much weaker with the collapsed conformations of TF than with the extended conformation , presumably due to the lack of Arm1’s availability in the former . This highlights the importance of Arm1 in TF’s interactions with unfolded substrate proteins . While our aim is to provide a molecular view on TF’s interactions during progressively more structured stages of protein folding , the partial folds in our simulation might also be seen as individual proteins . A comparison between the different systems then shows that the N-terminal domain plays an important role in the early interaction stages for all proteins , whereas the precise interaction maps depend on the specific details . Interactions with less structured and more hydrophobic substrates are clearly stronger and more stable . Our results provide the first view of the atomistic underpinnings of how a chaperone is able to interact with a protein chain during folding and to guide it to its native state . Taken together , they reveal a model of a highly adaptive chaperone , in which TF grabs the partial folds with its domain tips as they form , transfers them to its cradle to encapsulate them so that they are protected against misfolding interactions . TF , thus , acts as a “cradle for partial folds” that uses its flexibility to adjust its structure to the protein substrates of diverse size and composition , and protects the non-folded and partially folded chains against misfolding interactions with other domains or other proteins . It is worth noting that this encapsulation mechanism is distinct from that of chaperones that require ATP for their functioning , such as GroEL/GroES . TF does not require an ATP cycle to enable both substrate encapsulation and liberation in chaperones , and employs its own structural flexibility to achieve the same . This model provides a structural explanation for the ability of TF to interact with and stabilize a wide spectrum of substrates , ranging from unfolded chains to the diverse intermediate folded structures en route to the native state [18] .
Table 1 lists the various systems that were simulated . The systems involve the native structures of full Maltose Binding Protein ( MBP ) ( PDB code: 1JW4 ) , as well as two of its folding intermediates or partial folds—named P1 ( residues 7–62 ) and P2 ( residues 113–248 ) . Their structures were acquired from the PDB file for full MBP by removing the other residues . We employed Gromacs 4 . 5 ( or 4 . 6 in some cases ) [21] to perform MD simulations of these protein substrates in isolation , as well as with the trigger factor ( PDB code: 1W26 ) in several different starting configurations . The atomic interactions were defined by the Amber03 [22] force field for most systems . In a small set of recently concluded simulations , we used the Amber99SB-ILDN [23] force field in light of the findings of Beuchamp , et al . [24] . To all PDB structures , hydrogen atoms were added for the protonation state at pH 7 . 0 . After a steepest descent energy-minimization we solvated the structures in a dodecahedron periodic box ( of diameter 15–16 nm ) with TIP/3P water molecules for the Amber03 systems , and in TIP/4P water molecules for the Amber99SB-ILDN structures . The box sizes were set to allow for sufficient translational and rotational dynamics of both molecules . The total systems size varied from 11356 atoms for the folded P1 only to almost 500000 atoms for the extended TF with an unfolded P1 . The systems were neutralized by adding Na+ and Cl− ions ( 50 mM NaCl ) . Energy-minimization , followed by 50–100 ps position-restrained ( all heavy atoms of protein restrained with force constant of 1000 N/m in each direction ) MD runs resulted in equilibrated positions of water molecules and ions . The systems were coupled with a v-rescale thermostat ( τ = 0 . 2 ) and a Parrinello-Rahman barostat ( τ = 1 . 0 and reference pressure = 1 Bar ) . The bonds in the protein were constrained using the LINCS algorithm [25] . Coulombic and van der Waals interactions were treated with a cut-off radius of 1 . 0 nm , while long-range electrostatic interactions were handled using the PME algorithm ( with a mesh size of 0 . 12 nm ) . Parallel production runs were performed ( with different random starting velocities ) for 200–300 ns at close to room temperature ( 300 K ) with a time step of 2 . 0 fs ( coordinates were written every 2 . 0 ps ) . Each 200–300 ns long MD simulation trajectory ( started with varying random velocities ) took on the order 50000 CPU hours . Four runs were performed on a TF-MBP system , in which the MBP’s crystal structure was placed at a center of mass distance of 8 nm ( with minimum distance of approx . 3 nm ) from the extended structure of TF . For the set of six runs on TF-P2 complex , the P2 was placed at a center of mass distance of 7 nm ( with minimum distance of nearly 2 . 5 nm ) from the extended TF . P1 , with 56 residues the smallest of the five partial folds , was studied using several initial conditions of both P1 and TF: three sets of runs for the folded P1 with different starting configurations of TF , and three sets of runs for an unfolded P1 chain with different initial configurations for TF . In the former set of simulations , P1 was placed at a minimum distance of 2 . 3–3 . 5 nm from: the extended TF structure ( PDB structure ) , the semi-collapsed ( SC-TF ) conformation , and the fully-collapsed ( FC-TF ) conformation ( last two obtained from isolated TF simulations [12] ) . The P1 complex with extended TF was named “Touching Complex . ” Similarly , the three sets performed on TF-unfolded P1 systems were also initiated with the unfolded P1 at a minimum distance of 3–4 nm from the three conformations of TF—extended , semi-collapsed , and fully-collapsed . In all TF-P1 systems , four to eight simulations were run for 300 ns , resulting in the total aggregate time of about 15 μs . The first 200–300 ns provide an overview of the early interactions between TF and its substrates . However , to further capture the post-folding dynamics of P1’s interactions with TF , we bypassed the spontaneous but very slow transition towards complete binding of the substrate to TF . To this end , we used an energy minimized configuration of the TF–P1 complex as an initial conformation to perform additional MD simulations , and acquire insight in the more strongly bound states . To provide such a minimum energy conformation , we performed molecular docking calculations on the TF–P1 complex . Docking calculations with the Rosetta Docking server [26–28] gave 10 minimum energy configurations . Starting from the three lowest energy configurations , we performed three additional 100 ns MD simulations to evaluate the stability of these complexes , called “Hugging Complexes” . Two of the final conformations that we obtained were used as starting structures for pulling experiments . We performed steered MD ( SMD ) calculations on P1 in isolation and two TF-P1 systems by pulling at the ends of the P1 ( residues 7 and 62 , respectively ) . PLUMED 2 . 0 [29] was employed in combination with GROMACS 4 . 6 and the Amber99SB-ILDN force field [23] . The systems were placed in a dodecahedral water box of 15 nm ( and 50mM NaCl ) and output frequency of 100 time steps ( with time step of 2 . 0 fs ) . We attached dummy beads at the two ends through a harmonic spring ( with a spring constant of 50 kN/nm ) , and pulled them apart at a speed of 0 . 5 nm/ns , unfolding P1 in about 30 ns . To prevent jumps in the pulling coordinate when the distance to a periodic image would be shorter , interactions between periodic images were switched off for the pulling coordinate ( although , for all other interactions , periodic boundary conditions are taken into account ) . Initial conditions for the SMD runs were obtained from 100 ns AA-MD simulations on the lowest energy Rosetta docked TF-P1 complexes , HC1 and HC2 respectively . All systems were prepared/equilibrated as described above . For each system , we ran 20 steered MD simulations , and obtained the work-extension curves . We computed the average work required for unfolding P1 in each system ( 〈W0−>t〉 ) , and by applying Jarzynski equality [30] , the potential of mean force ( PMF ) from the exponential average work ( kBT log〈exp ( −βW0−>t〉 ) . As the PMF is dominated by small values of W0−>t [31] a better approach is to use the stiff-spring approximation and compute the potential of mean force ( PMF ) from the first and second cumulant ( 〈W〉 − 0 . 5β ( 〈W2〉−〈W〉2 ) ) [31] . Note that the second cumulant in the PMF expression is the estimation of variance in work over the extension . Large fluctuations in the work-extension curves , thus , result in spuriously large negative values of PMF , rendering the PMF unreliable . To analyze the interactions between TF and substrate , we define the formation of a contact when the minimum distance between the heavy atoms of the two molecules falls below 0 . 4 nm . Such an event is labeled an attachment event . In contrast , a detachment event was recorded when , after the formation of a contact , the corresponding minimum distance increased above 1 . 0 nm . The margin prevents registration of false detachment events . The mean first binding time of the TF-substrate interaction measures the time , averaged over the number of trajectories , it takes to form a metastable complex after the first contact , i . e . , the last attachment event that is not followed by detachment events in rest of the MD simulations . The probability of a domain or residue of TF to form a contact with substrates ( the contact probability ) is estimated by determining the area of the normalized probability distribution of the minimum distance between the corresponding heavy atoms on TF and the substrate for values less than 0 . 4 nm . The number of partial fold contacts , or PF-contacts , in P1 are defined as the number of backbone hydrogen bonds between the β-strands formed by first five ( residues 7–11 ) and last five residues ( residues 58–62 ) of P1 . The fraction of hydrophilic contacts was calculated by dividing the number of contacts between nitrogen and oxygen atoms of molecules by total number of contacts between the respective molecules . Hydrophobic maps of molecules were calculated by the novel hydrophobic probes method developed by Acharya et al . , [32] as described in an earlier work [12] . The hydrophobicity map shows the location and the strength of the hydrophobic patches ( or residues ) on the TF’s surface . All analysis was performed using indigenous scripts and the built-in tools of GROMACS 4 . 5 and 4 . 6 [21] . VMD 1 . 9 was used for visualization of protein structures and creating images [33] . We concatenated the trajectories of the dynamics of unfolded P1 in isolation and in complex with extended TF , and performed principal component analysis on the dihedral angles of P1 [34] using GROMACS tools . After sorting the common eigenvectors in the descending order of the associated eigenvalues , we projected the concatenated trajectory along the first 15 eigenvectors . The 2-D projection along any pair of eigenvectors was obtained by the plotting the corresponding projections against one another . In this projection , the two sets of simulations ( P1 in isolation and P1 in complex with extended TF ) can be easily separated to differentiate the conformational spaces explored by them along common eigenvectors . Results of this analysis are discussed in the Supporting Information .
We performed multiple MD simulations of unfolded P1 with TF , which is initially either in extended ( crystal structure ) , semi-collapsed or fully collapsed conformation ( see Methods ) [12] . The unfolded chain attaches to the TF in almost all MD simulations , though the binding is initially transient , with multiple attachment-detachment events per binding site ( see S1 and S2 Figs ) . The number of these events , and the mean first binding time—defined as the average time after which no detachment events are observed ( see Methods ) —are the smallest for the unfolded P1-extended TF system ( Table 2 ) . The two form a stable complex in a very short time , after detaching only once . These values increase significantly for the semi- and fully-collapsed conformations of TF . This increase can be easily understood , as in the latter two conformations , Arm1 forms a contact with PPIase domain ( with an additional interface between N-terminal and Arm2 for fully-collapsed TF ) , hence can no longer bind to the substrate . In that case , while TF can still bind the substrate proteins , the interactions may be diminished . This is in agreement with Thomas et al . [13] who , building up on the work of Gupta et al . [35] , proposed that the formation of an interface between PPIase domain and Arm1 in isolated TF might affect the interactions between TF and nascent chains . Fig 2A plots the contact probability for TF’s residues with unfolded P1 ( In this work we represent the tendency of interaction by contact probability rather than the interaction free energy , because of the non-equilibrium nature of the simulations ) . The bottom-most panel shows the root-mean square fluctuations of the C-α atoms of TF residues . The extended TF’s interactions with unfolded P1 are non-specific and distributed over the entire surface of TF , with strong interaction sites at the N-terminal ( residues 38–42 ) , Head ( residues 184–186 and 192–194 ) , Arm1 ( residues 320–325 ) and Arm2 ( residues 378–390 ) . The strongest contacts are made at hydrophobic residues , viz . , Phenylalanine , Isoleucine and Tyrosine . The flexible tip of Arm1 , which also contains a strongly hydrophobic patch , hosts the strongest interaction site . Because the PPIase domain is bound to the Arm1 in the semi-collapsed and fully-collapsed conformations of TF , both these domains are unavailable for substrate binding . Consequently , there are no interaction sites on PPIase domain and Arm1 in the SC and FC conformations . They are replaced by new sites on Arm2 ( in particular , Lys368 ) and HA1-Linker ( residues 254–275 ) in P1’s association with semi-collapsed TF . In case of the fully-collapsed TF , the mostly hydrophilic external surface of N-terminal dominates the interaction sites with the substrate . Fig 2B maps the contact probability on an extended TF structure . Most of the interaction sites ( blue ) lie at the tips of various domains , except for those on the Head , which covers the attached chain with its inner surface , like a lid . Our results corroborate the findings of Lakshmipathy et al . [36 , 37] and Saio et al . [14] , who showed that TF employs both arms—especially the flexible loops at the tips of the arms—and N-terminal to bind nascent polypeptide chains . Fig 2A compares the interaction sites estimated in this work with those reported by Saio , et al . [14] . The interacting residues identified by Saio et al . are represented by black squares on the x-axis of top panel ( Please note that Saio et al . did not provide information on the relative strengths of these interaction sites ) . Almost all of the interaction sites identified in their work coincide with the peaks of the contact probability map found in our simulations , especially at the tips of the arms and inner surface of Head . However , our simulations did not capture the relatively long process of TF fully encapsulating the unfolded substrates , and hence , the interaction sites in the cleft of the cradle—between residues 350 and 370 . Instead , our simulations highlight the events in the early interaction dynamics , where TF binds unfolded substrates with the tips of its appendages ( including N-terminal ) , and subsequently collapses around them to strengthen the interactions . To investigate how the early interactions with TF affect the unfolded P1 , we compared its dynamics in complex with extended TF with that of unfolded P1 in isolation . We focused on the effect of TF on the ends of P1 , which are in a strong direct contact with each other in the folded state . We observed that in the presence as well as absence of TF , the unfolded P1 collapses into a molten globule with a small radius of gyration , comparable to that of folded P1 . This spontaneous collapse brings P1’s ends together , leading to the formation of strong transient non-native contacts , which can result in misfolded states . TF slows down this spontaneous collapse and largely prevents the formation of non-native end-to-end contacts in P1 . This is evident in the distribution of the number of P1’s end-to-end contacts , plotted in the left panel of Fig 2D . The presence of TF shifts this distribution to lower values w . r . t . the unfolded P1 in isolation , resulting in almost zero end-to-end contacts in P1 . In contrast , the distribution for folded P1 is narrower than unfolded P1 and peaks at a higher number of contacts . The right panel of Fig 2D shows that the distribution of the end-to-end distance in P1 in complex with extended TF is wider and shifted to larger distance w . r . t . P1 in isolation . The distribution for folded P1 in isolation has a sharp peak at an even lower value of 0 . 5 nm . An increase in the end-to-end distance in presence of TF translates into a higher radius of gyration of P1 ( see Fig 2E ) . The unbound P1 in isolation and in the system with TF have comparable compactness . Upon association , the distribution shifts to higher radius , implying that TF not only hinders collapse , but also stretches collapsed P1 , likely avoiding bad intramolecular contacts and facilitating sampling of larger conformational space ( see S2 Fig ) . Fig 2F plots the minimum distance TF and P1’s ends versus the radius of gyration of P1 for all seven trajectories that demonstrate binding . Unfolded P1 can either first collapse and then bind TF ( left panel ) , or vice-versa ( right panel ) . Strikingly , in both cases P1 opens up after associating with TF , and samples higher values of radius of gyration . The highest values are sampled in the trajectories where binding occurs before the collapse . One of the final TF-P1 complexes , shown in Fig 2C , demonstrates a relatively open conformation of P1 , held by the Head and the tips of the arms . The deformed conformation of TF suggests that extended TF adapts to the substrate size and structure ( see S2 Fig ) . These results point to a mechanism for the chaperone action of TF , wherein TF: 1 ) binds to an unfolded chain; 2 ) prevents the formation of spontaneous transient , non-native end-to-end contacts; 3 ) slows down the chains collapse; and , 4 ) facilitates the chain to explore more conformations than the unfolded chain can in isolation . Mashaghi et al . found recently that TF binds partially folded structures formed in MBP chains [18] . In order to elucidate the structural basis for this binding , we performed MD simulations of folded P1 in the presence and absence of TF ( starting in the extended conformation ) . We found that folded P1 indeed forms a complex with TF , though this requires more time than for unfolded P1 , and involves more frequent detachments ( Table 2 ) . Semi-collapsed and fully-collapsed conformations of TF display shorter mean first binding times and fewer detachment events , even slightly smaller compared to unfolded P1 . This may be because these TF conformations preferably bind to the near-folded substrates or folding intermediates in cytosol to isolate them . The top panel ( TC ) of Fig 3A shows contact probabilities of TF residues with folded P1 in the extended TF simulations . Most of the interaction sites lie on the largely hydrophilic N-terminal ( residues 24–32 and 45–54 ) , and only a few on Head and hydrophobic Arm1 , indicating that folded P1 explores fewer sites than the unfolded P1 ( Fig 2A ) . The corresponding contact probabilities of the P1 residues are plotted in the top panel of Fig 3B . The interaction sites are uniformly distributed over the entire surface of P1 , including the ends . For extended TF , most of the interaction sites ( blue ) are located on the flexible loops at the tips of N-terminal ( residues 30–50 ) and Arm1 ( Fig 3E ) . Similar analysis ( see S3 Fig ) for the semi-collapsed and fully-collapsed conformations of TF reveals that folded P1 attaches primarily to the N-terminal , and that the other domains rarely interact with folded P1 . Thus , TF initially interacts with folded P1 via the tips of one or two of its appendages , dominated by N-terminal . We therefore refer to these complexes as “Touching Complexes” ( TC ) . With the aim to probe the post-folding dynamics of P1’s interactions with TF , we initiated MD simulations using the lowest energy configurations for rigid TF and P1 obtained from docking calculations ( see Methods ) . We found that in these simulations , P1 moved closer to the cradle center , and was encapsulated by the flexible TF appendages , led by Arm1 . Therefore , we referred to these structures as “Hugging Complexes” ( HC ) ( see S4 Fig ) . The contact probabilities for the P1 residues in these simulations ( plotted in the lower three panels in Fig 3B ) are less uniform than those in the TC . The corresponding three panels for TF in Fig 3A show strong interaction sites with Arm1 ( yellow ) for all HCs . These interaction sites also occur in TC , indicating the central role of Arm1 in substrate binding . P1 typically interacts with two TF elements: either with Arm1 and PPIase , or with Arm1 and the N-terminal . Fig 3G shows a representative final HC conformation ( HC2 ) , where P1 binds at the Arm1 and PPIase domain . This complex is similar to the final TC conformation shown in Fig 3F , suggesting an ( partial ) overlap between the conformational space of the TC and HC simulations . To investigate whether TF-binding stabilizes P1 , we monitored the number of intact ( out of four ) β-sheet hydrogen bonds between the strands formed by residues 7–11 and 58–62 , referred to as “partial fold” ( or PF ) contacts . Probability distributions for these contacts are plotted in Fig 3C . In the full MBP , these hydrogen bond contacts are almost all intact , stabilized by the surrounding structure of MBP . In contrast , the truncated P1 in isolation is less stable and shows loss of PF-contacts , with even a ( small ) peak at zero contacts . The binding of P1 with TF in Touching Complexes stabilizes it w . r . t . P1 in isolation , as illustrated by the loss of the peak at zero contacts . Hugging Complexes show more intact contacts than the TC . This is consistent with the increased proximity of P1 to the TF’s cradle in the Hugging Complex , as indicated by the corresponding probability distributions of the center of mass ( COM ) distance between two molecules in Fig 3D ( dashed lines ) . The probability distribution for the number of TF-P1 contacts shown in the same figure ( solid lines ) , indicates that the increased proximity results in a higher number of TF-P1 contacts . Thus , after or during folding , P1 can move towards TF’s interior , which increases the stabilization of its PF-contacts . This mechanism is reminiscent of the “cradle model” for TF’s function , which speculates that the protein folding and stabilization happen in the cradle-like interior of TF [10] . A comparison between the unfolding potential of mean force ( PMF ) of folded P1 in isolation using steered MD ( SMD ) , and that of P1 in Hugging Complexes ( S5 Fig ) provides more insight into the nature of TF’s stabilization of P1 . As the interactions maps in the HCs 2 and 3 are comparable , we selected the final configuration of the MD simulation of HCs 1 and 2 as input for the SMD . Pulling at the ends ( first and last residues ) of folded P1 , we measured the cumulative work required to unfold P1 ( see Fig 4A ) . We found that , compared with P1 in isolation , a higher amount of work is needed to unfold P1 in complex with TF . Fig 4B indicates that the loss of P1’s β–sheet network is the primary contributing factor in the free energy barriers in the pulling process . The β–sheet network is formed by two sets of backbone hydrogen bonds: those between residues 7–11 and 58–62 , and those between residues 7–11 and 35–39 . There are , on average , four hydrogen bonds between each pair of β–strands , and the sudden loss of each set coincides with a jump in work . The changes in TF-P1 contacts ( Fig 4C ) do not seem to contribute to the free energy barriers . We used Jarzynski’s inequality [30] , and the second cumulant approximation [31 , 38] to turn the non-equilibrium work-extension curves into equilibrium free energy . In the early unfolding regime—extension of 0 . 7 nm over 1 . 4 ns of pulling , the exponential average of work and the second order cumulant expansion coincide ( Fig 4D ) , which confirms the reliability of PMF in this regime [31 , 38] . For the complete unfolding process , however , the variance in the work-extension graphs was too large , making the resulting PMF unreliable . For this reason , we focused only on the early unfolding regime , which contains the first major free energy barrier in all three systems , caused by the loss of PF-contacts ( shown by circles in Fig 4E ) . The energy barriers in both the TF-P1 systems are significantly bigger than that for P1 in isolation . Breaking these PF-contacts , thus , requires more energy when P1 is in complex with TF than in isolation . The number of TF-P1 contacts ( solid lines ) decrease by ∼ 10 in both HC1 and HC2 . This loss may also contribute to the amplification of the free energy barrier . Fig 4F visualizes the loss of PF-contacts through breaking of hydrogen bonds between the first and last β–strands ( colored orange ) upon pulling . To summarize , TF stabilizes protein’s partial folds by encapsulating them in its interior ( and adapting to their structure ) , and strengthening their PF-contacts . Once the smallest partial-fold ( P1 ) is completely folded , TF further interacts with the larger partial folds ( e . g . , P2 ) , eventually interacting with full MBP . We simulated these interaction through six MD runs of TF with a larger partial fold P2 , and four runs with full native MBP molecule ( see S6 and S7 Figs ) . The contact probability plots in Fig 5A ( top panel for MBP , bottom panel for P2 ) show that for both systems , the tip of N-terminal domain ( residues 38–55 ) is responsible for the majority of the interactions , while other domains play only minor roles . P2 also binds to the tips of Arm1 ( residues 320–325 ) . The TF structures in Fig 5B visualize the location of these interaction sites with MBP and P2 . Most of them lie at the tips of TF’s appendages , while the PPIase domain again appears to cover P2 , like a lid , with its inner surface . The hydrophobicity maps show that the dominant interaction sites for both substrates lie in hydrophilic regions . Fig 5D visualizes the TF-substrate binding with two representations of TF-MBP and TF-P2 complexes . Both substrates demonstrated association and dissociation events with TF in all simulations . The mean first binding time as well as the number of detachment events for P2’s binding with TF were comparable with those of P1’s binding with extended TF ( Table 2 ) . Native MBP , on the other hand , detaches more often over a longer duration before it forms an apparently metastable complex with TF . The stability of TF-substrate complexes increases for substrates with lower degree of structure , i . e . , native MBP interacts weakly with TF relative to the partial folds , which in turn form complexes less readily than the unfolded substrates . These findings match the observations of Mashaghi et al . [18] , who showed that native MBP does not bind to TF as strongly as its partial folds do . This suggests that as the MBP folds , its binding with TF progressively weakens , which can help in the post-folding release of the protein from TF . To investigate the variation in the nature of TF’s interactions with different partial folds , we compared the ratio of hydrophilic contacts to total contacts , between TF and different substrates ( Please note that this estimate has a bias towards a large number of mixed contacts , i . e . , Oxygen and Nitrogen atoms with Carbon atoms , and is therefore conservative ) . Fig 5C plots the distribution of these ratio in the first 10 ns of contact ( dashed lines ) and during the entire duration of contact ( solid lines with circles ) . Overall , the TF-MBP interactions are most hydrophilic , followed by extended TF-unfolded P1 and TF-P2 interactions , with extended TF-folded P1 interactions being the most hydrophobic . The small positive shifts in this ratio for folded P1 and P2 after the first 10 ns indicates that for partial folds , hydrophobic contacts initiate the binding , which becomes gradually more hydrophilic . In case of unfolded P1 , the distribution in the first 10 ns is wider and lower , while the overall distribution has a distinct sharp peak comparable to full MBP . This suggests that the large available surface area in unfolded P1 enables both hydrophobic and hydrophilic contacts to initiate the binding with TF . In a recent study , the chaperone Trigger Factor was shown to direct Maltose Binding Protein ( MBP ) chains to their native state by transiently stabilizing specific partially folded intermediate states [18] . These findings allowed us to undertake an all-atom molecular dynamics simulation study of the interaction between Trigger Factor and the MBP protein chain in its different stages of folding . We found that TF interacts with folded structures with the ends of its flexible appendages , especially N-terminal , thus forming a “Touching Complex” . The residues 45–54 ( tip of N-terminal ) and 24–32 ( also N-terminal ) interact most frequently with the substrates . During the interaction with the substrates , TF’s secondary structure remains intact , while its tertiary structure can adapt to the substrates , as indicated by structural deviations in the flexible linker regions ( see Supporting Information ) . TF can leverage this flexibility to wrap its appendages around the folded structures ( P1 ) , forming the more stable “Hugging Complexes , ” while other MBP segments remain unfolded . Our steered MD simulations showed that breaking the end-to-end β–sheet hydrogen bonds of P1 in these complexes requires more work than when P1 is in isolation . This suggests that by wrapping around partially folded structures and establishing multiple favorable contacts , TF stabilizes these partial folds . The flexible appendages and the heterogeneous nature of its surface allow TF to bind to different intermediate folds of the same protein , and , by induction , indeed to proteins of different sizes . This binding is the strongest with unfolded protein chains , and weakens gradually with the progressive folding of the polypeptide . It weakens significantly upon the formation of fully folded MBP , which likely plays a role in the post-folding release of the protein . The unfolded chain binds simultaneously with two to three sites on TF , in particular with the flexible loops at the tips of N-terminal ( residues 38–40 ) , Arm1 ( residues 320–325 ) and Arm2 ( residues 378–380 and 387–390 ) . In TF’s interactions with nascent polypeptide chains , the tips of the arms were found to be important binding sites in the NMR experiments by Saio , et al . [14] , as well as the fluorescence spectroscopy measurements by Lakshmipathy et al . [36 , 37] . This consistency validates the relevance of interactions predicted by our extensive molecular simulations . It has been observed that the Arm1 plays an important role in TF’s collapse by strongly interacting with the PPIase domain [12 , 13] . Consequently , Thomas , et al . speculated that the TF’s collapse may adversely affect its interactions with unfolded chains . Our findings confirm this suggestion—the binding of unfolded polypeptides is indeed much weaker with the collapsed conformations of TF than with the extended conformation , presumably due to the lack of Arm1’s availability in the former . We found that unbound to TF , unfolded P1 collapses spontaneously into a molten globule with the formation of strong non-native contacts , which can lead to misfolding . In contrast , TF slows down , and even reverses , this collapse , thus allowing the unfolded chain to explore a larger conformational space than in isolation . We can , thus , attribute TF’s role in protein folding to creating kinetic traps in the formation and , subsequently , breaking of the intramolecular contacts of the protein’s folding intermediates in order to prevent misfolding and aggregation . Employing coarse-grained MD on TF-ribosome complex , O’Brien et al . have suggested a similar chaperone action for TF at the ribosome [19] . Together with the previous findings , our results suggest the following scenario , which is summarized in Fig 6 . TF employs the flexible loop at its domain tips ( mostly its arms ) to bind a small part of the MBP chain . As the TF transfers this part of the chain into its cradle , the chain forms ( either simultaneously or subsequently ) a partial fold . TF then shields and isolates this partial fold from the rest of the protein , and continues to fold and stabilize intermediates of increasing size until the native state of MBP is reached . The weak binding of native MBP with TF facilitates its subsequent release into the solution . In light of the recent work by Saio , el al . [14] , it is likely that more than one TF monomer may be involved in subsequent folding of MBP . This model provides a structural explanation for the ability of highly adaptable TF chaperone to interact with and stabilize a wide spectrum of substrates , ranging from unfolded chains to the diverse intermediate folded structures en route to the native state [18] . By providing a flexible encapsulating cradle , TF protects unfolded and partially folded chains against misfolding interactions with other domains of MBP or other proteins . We realize that the use of truncates and short trajectories cannot represent the full folding , binding and chaperone action , and that atomistic force fields are not perfect . Nevertheless they can give additional atomistic insight . Among the different atomistic force fields Amber03 and Amber99SB-ILDN , especially the latter , are perceived to be reasonably good for protein folding and unfolding transitions [24] . However , a full comparison between different force fields is beyond the scope of this study . Regardless , this work provides the first atomistic view on a chaperone-client complex in different stages of folding , offers an ( atomistic ) explanation for the ability of TF to guide chains to their native state by stabilizing partial folds and protecting them against misfolding and aggregation interactions . It builds on the previous work by Saio , el al . [14] and O’Brien et al . [19] , and confirms that TF’s flexibility is central to its ability to interact with a wide range of client states . While it lends a new hypothesis about the Trigger Factor MBP system in particular , it also yields generic insights into how an ATP-independent chaperone can assist folding and prevent misfolding . There are many questions that still need to be explored . For example , how does TF distinguish partially folded structures from fully folded structures ? It is not clear how precisely the substrate transitions from one state to the next: does TF completely detach or remain attached as the substrate progressively folds ? Do other chaperones exploit similar mechanisms ? We expect advanced molecular simulation methods , along with increasing computer power , to complement improving experimental techniques in answering these questions in a very near future .
|
Trigger Factor ( TF ) is an ATP-independent chaperone protein that assists in folding and prevents misfolding . Up to now , it is a general unsolved question how chaperones assist in the folding of protein chains . Experimental methods that can probe at the length and timescales of inter-residue interactions are scarce , while the systems are too large—and the folding process too long—to be studied by computer simulations . To overcome these obstacles , the authors performed molecular dynamics simulations at key moments along the folding pathway , and address the changes in the folding and unfolding dynamics of protein chains while in contact with TF . This study provides the first detailed view on a chaperone-protein complex in different stages of folding and offers an explanation for the ability of TF to guide chains to their native state . Moreover , the results demonstrates the role of TF’s flexibility in interacting with a wide range of client states . Overall , it explains how TF can interact with many types of substrates in various stages of folding , without the need for an ATP cycle to switch between encapsulation and liberation of client proteins .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"and",
"Discussion"
] |
[] |
2015
|
The Trigger Factor Chaperone Encapsulates and Stabilizes Partial Folds of Substrate Proteins
|
Despite the availability of vaccines , influenza remains a major public health challenge . A key reason is the virus capacity for immune escape: ongoing evolution allows the continual circulation of seasonal influenza , while novel influenza viruses invade the human population to cause a pandemic every few decades . Current vaccines have to be updated continually to keep up to date with this antigenic change , but emerging ‘universal’ vaccines—targeting more conserved components of the influenza virus—offer the potential to act across all influenza A strains and subtypes . Influenza vaccination programmes around the world are steadily increasing in their population coverage . In future , how might intensive , routine immunization with novel vaccines compare against similar mass programmes utilizing conventional vaccines ? Specifically , how might novel and conventional vaccines compare , in terms of cumulative incidence and rates of antigenic evolution of seasonal influenza ? What are their potential implications for the impact of pandemic emergence ? Here we present a new mathematical model , capturing both transmission dynamics and antigenic evolution of influenza in a simple framework , to explore these questions . We find that , even when matched by per-dose efficacy , universal vaccines could dampen population-level transmission over several seasons to a greater extent than conventional vaccines . Moreover , by lowering opportunities for cross-protective immunity in the population , conventional vaccines could allow the increased spread of a novel pandemic strain . Conversely , universal vaccines could mitigate both seasonal and pandemic spread . However , where it is not possible to maintain annual , intensive vaccination coverage , the duration and breadth of immunity raised by universal vaccines are critical determinants of their performance relative to conventional vaccines . In future , conventional and novel vaccines are likely to play complementary roles in vaccination strategies against influenza: in this context , our results suggest important characteristics to monitor during the clinical development of emerging vaccine technologies .
Seasonal and pandemic influenza pose major public health challenges [1 , 2] Vaccines against seasonal influenza aim to raise antibodies against the hemagglutinin ( HA ) and neuraminidase ( NA ) surface proteins of circulating strains [3] . While these targets offer the best correlates for immune protection , they are also by far the most variable amongst influenza viral components [4 , 5] , undergoing continual evolution for immune escape: current seasonal influenza vaccines therefore need to be updated regularly . Moreover , influenza pandemics are caused by the emergence of a virus with an altogether new HA ( and other viral components ) , to which there is little or no immunity in the human population [6 , 7] . Current vaccines cannot be deployed in advance of an influenza pandemic , as it is not possible to predict what virus will cause the next pandemic [8] . There is evidence to suggest that other viral components , including the matrix protein M1 and the nucleoprotein NP , may be more conserved than HA and NA [9–13] . Immunity to these proteins , mediated by T-cells rather than by antibodies , is associated with broad-spectrum protection[14] , even against novel pandemic viruses [15 , 16] . At the same time , it is also possible for antibodies to raise broad-spectrum protection: with most of HA variability concentrated in the ‘head’ region of the protein , antibodies against the more conserved ( but less accessible ) ‘stem’ region have also attracted considerable attention[17–19] . Antibodies against the ion channel protein M2 have also been shown to elicit broad protection [20 , 21] . Only with recent advances in vaccine technology has it been possible to target these alternative viral components [10 , 17–19 , 22–24] . The resulting emergence of candidates for ‘universal’ vaccines raises the potential for more stable influenza vaccination programmes , that do not have to be updated so frequently . At the same time , even with current , strain-matched vaccines , population coverage is on an increasing trend: in some settings ( notably in the UK ) there is growing emphasis on widened vaccination coverage to reduce transmission as well as disease [25] . Coverage in the US has been steadily rising and has recently exceeded 43% of the population [26] . These trends suggest that annual , mass influenza immunization programmes could foreseeably become a reality . Together , such developments raise important questions about the potential future use and impact of influenza vaccines . For example , how might novel vaccines compare against current , strain-matched vaccines , in their ability to control transmission ? What are the implications for seasonal HA evolution , of a mass immunisation programme targeting HA versus one targeting other more conserved viral components ? As these vaccines are still in development , important vaccine parameters , including classical vaccine efficacy , and duration of protection in humans , remain to be determined [27]: what are the implications of these vaccine characteristics , for future immunization programmes ? Previous work [28] focused on the emergence of a pandemic virus , finding that the ability of cross-protective vaccines to mitigate pandemic risk depended on the ability of any vaccine ( whether current or future ) to provide broader protection than that provided by natural infection . Another modeling study [29] showed how cross-protective vaccines could slow the rate of antigenic evolution for seasonal viruses , thus enhancing the control of seasonal epidemics with conventional ( HA-specific ) vaccines . However , neither model addressed the potential effect of conventional vaccines on seasonal viral evolution , and how this might compare with universal vaccines . The present model builds on this previous work , addressing the questions above with a simple , novel model of influenza transmission and evolution . The model evaluates the relative merits of ‘conventional’ versus ‘universal’ vaccination , while casting light on vaccine characteristics that would be helpful to quantify , in anticipation of novel vaccine candidates entering advanced clinical trials .
The epidemic component is a deterministic , compartmental framework that models each seasonal epidemic as a single epidemic wave , with a single circulating strain ( Fig 1A ) . For simplicity we ignore age structure , as well as spatial heterogeneities , assuming simply a fully ‘well-mixed’ population . The governing equations are as follows: dSdt=−λSdIdt=λS−γIdS ( cp ) dt=−λS ( cp ) dI ( cp ) dt=λS ( cp ) −γI ( cp ) dR ( cp ) dt=γ ( I+I ( cp ) ) ( 1 ) Here S , I , R are respectively the proportions of the population who are susceptible to infection; infectious; and recovered and immune . The superscript ( cp ) marks individuals having cross-protective immunity ( but not strain-specific immunity ) ; γ is the per-capita rate of recovery; and λ is the force of infection , given by: λ=βI+ ( 1−c ) βI ( cp ) Here β is the effective contact rate , multiplied by the average number of infections per infected case , and c is the reduction in infectiousness arising from cross-protective immunity , written so that c = 1 corresponds to fully transmission-blocking immunity . The initial conditions are given by the proportion of the population that has HA-specific immunity , given prior epidemic sizes and the amount of antigenic drift that has occurred ( see below ) , along with two types of vaccination programme , which are completed prior to each epidemic and with random coverage , irrespective of an individual’s exposure history or immune status: conventional vaccination displaces individuals from S to R and from S ( cp ) to R ( cp ) , while universal vaccination displaces individuals from S to S ( cp ) and R to R ( cp ) . For comparability between the two types of vaccine being considered here , it is necessary to choose values for the quality of vaccine protection ( vaccine ‘efficacy’ ) that are matched in terms of their population effect . We assume for simplicity that universal vaccination elicits the same cross-protective immunity as does natural infection , thus identifying c with the efficacy of universal vaccination . Correspondingly for conventional vaccines , we assume that efficacy derives from a proportion c of vaccinated individuals successfully acquiring strain-specific immunity ( the rest remaining with their prior immune status ) . It is straightforward to show ( see appendix ) that both vaccines thus have the same effect on R0 . The initial conditions are given by the proportion of the population that has HA-specific immunity , given prior epidemic sizes and the amount of antigenic drift that has since occurred ( see below ) , along with two types of vaccination programme , which are completed prior to each epidemic and with random coverage: conventional vaccination displaces individuals from S to R and from S ( cp ) to R ( cp ) , while universal vaccination displaces individuals from S to S ( cp ) and R to R ( cp ) . In the interepidemic period , we model a loss of immunity in the population due to three mechanisms: loss of strain-specific immunity through antigenic drift , loss of cross-protective immunity through waning of T-cell immunity , and a general depletion of immunity through population turnover ( replacement of immune hosts by susceptible ones ) . These are implemented as follows . For antigenic drift we adopt a simple deterministic framework to capture the essential role of population immunity , in driving selection for new variants ( see , for example , ref [40] for a review of evidence supporting this assumption ) . We assume a one-dimensional axis of HA antigenic variation , a simplified representation of the distinctive ladder-like phylogeny of influenza A hemagglutinin [41] . Fig 1B shows the simple case of a single immunizing strain ( a ‘reference’ strain ) . We denote d as the antigenic ‘distance’ between this and a candidate virus , shown on the horizontal axis . The Figure captures two essential features of antigenic evolution: first , candidates with greater d have a greater degree of immune escape and therefore a higher transmission potential [39] ( blue curve ) . However , they arise at a lower frequency ( red curve ) . Combining these two opposing factors to yield the ‘frequency-weighted immune escape’ ( orange curve ) , we assume that—on a population level—the selected virus for an upcoming season is one that maximizes this quantity . Specifically , the frequency-weighted immune escape for this candidate virus is defined as: F ( d ) =exp ( −kd ) [1−exp ( −d ) ] , ( 2 ) where k is a parameter governing the relative rarity of immune escape variants . For example , in the theoretical case k = 0 there is unlimited viral diversity in the interepidemic period , thus allowing a pandemic-scale outbreak every year . At the other extreme as k → ∞ , there is no generation of escape variants even in the face of population immunity: a situation similar to measles . For influenza , the scenario is intermediate . We calibrate the value of k in order to yield , at steady state , seasonal epidemics that infect roughly 10% of the population per season , consistent with the behaviour of seasonal influenza [42–44] . While eq ( 2 ) is in the simple case of a population with exposure to only one virus , over several seasons there is a series of viruses that emerge and circulate . Moreover , conventional vaccination in any given season offers protection against the virus circulating in that season , but also—to an extent diminishing with antigenic distance—against related viruses . It is thus necessary to keep track of the exposures to these viruses in the population , and to evaluate the proportion susceptible over all of these histories . Nonetheless , as we assume a one-dimensional antigenic space , it is only necessary to record the most recent infection or vaccination that individuals have undergone . Details of the necessary record-keeping are provided in the appendix . For the waning of T-cell immunity , we assume simply that a proportion σ of individuals lose this immunity in every interepidemic period . For illustration we choose σ = 0 . 21 , consistent with findings from early seminal work that suggested a T-cell half-life of 3 years [38] . However , it is important to note that there is considerable uncertainty around this Figure , with more recent studies suggesting that CD8 T-cell immunity can last as long as a decade , both for influenza ( [45] ) and for other viruses ( [46] ) . Accordingly , we explore this uncertainty in the work below . Table 1 shows the default parameter values used , and Fig 2 schematically summarises the procedure . Starting with a virus in a fully susceptible population , we simulate its spread using ( 1 ) ( ‘initiation’ in Fig 2 ) . We then simulate the selection for a new immune escape variant using ( 2 ) . Having determined this variant , we find the associated initial conditions ( population susceptibility ) for the subsequent epidemic season , and repeat the iteration from ( 1 ) to ( 2 ) ( ‘Circulation’ in Fig 2 ) . Finally , to study how a pandemic would be affected by the conditions of immunity in this population , at year 25 we introduce a virus to which only pre-existing cross-protective immunity , and not HA-immunity , is effective ( ‘Pandemic’ in Fig 2 ) . Although showing a steady state in Fig 2 , there are certain conditions where simulated seasonal influenza epidemics can show minor annual variations , as described below . Accordingly , we measure the ‘seasonal epidemic size’ as the mean epidemic size from years 5 to 24 . We additionally define the ‘pandemic size’ as the size of the pandemic when introduced at year 25 . The default parameter values shown in Table 1 ( second column ) are helpful for illustrating model behaviour . To examine the robustness of our model results to variation in these parameters , we then simulate the model through the range of plausible parameter values shown in Table 1 ( third column ) . In particular , using latin hypercube sampling , we generate 10 , 000 parameter sets within the ranges shown . To ensure plausible epidemiology , we retain those parameter combinations yielding seasonal epidemics that infect between 5 and 20% of the population , consistent with estimates that influenza infects roughly 10% of the population each season [45–47] . Under this parameter set , we then investigate the variability in the relative performance of conventional vs universal vaccines .
Fig 3 provides a side-by-side comparison of the effects of conventional and universal vaccines , presenting three different outcomes: control of seasonal epidemics ( panel A ) ; the effect of vaccines on the pace of antigenic evolution ( panel B ) ; and the implications of seasonal vaccination for pandemic sizes ( panel C ) . As described above , the Figure assumes equivalent vaccine efficacy and , in both cases , an annual vaccination program . The Figure is illustrative , involving only the point estimates for each of the input parameters involved ( Table 1 ) : below we examine the robustness of this qualitative behaviour under parameter variability . First , Fig 3A illustrates how conventional and universal vaccines could have differing effects on long-term patterns of influenza transmission . While both vaccines reduce seasonal epidemic sizes , at any given level of coverage , universal vaccines appear to have a stronger effect in suppressing seasonal epidemics . Moreover , Fig 3B illustrates—consistently with previous work—that large-scale universal vaccination would slow antigenic evolution over several seasons . Notably , however , these results suggest that conventional vaccines would tend to do the opposite , potentially accelerating antigenic change . We discuss below how these effects might arise from the different types of vaccine action . Under universal vaccination , the pace of antigenic evolution is driven to zero at sufficiently high coverage ( Fig 3B , orange curve ) : in this regime seasonal transmission is so heavily dampened that there is little strain-specific immunity to drive selection for new variants . However , we note that seasonal epidemics—even of very small sizes—could still occur at this coverage ( Fig 3A , ‘elbow’ in orange curve ) . This is a regime where universal vaccines interrupt transmission in the short term: over several years , however , seasonal viruses can sporadically persist , purely because of the accumulation of naïve individuals , rather than because of antigenic evolution—a situation analogous to measles ( [47] ) but with a substantially lower R0 . As discussed below , however , spatial and stochastic dynamics would greatly affect these extreme cases . Fig 3C additionally illustrates differences between the vaccines , for the size of a pandemic following several years of seasonal vaccination . The Figure illustrates that , while both types of vaccines can reduce seasonal epidemic sizes , high vaccination coverage with conventional vaccines tends to allow for increased pandemic sizes , whereas universal vaccines have the opposite effect . Moreover , pandemic sizes decline more rapidly with increasing universal vaccination coverage when there is no antigenic evolution ( i . e . an increased gradient in pandemic sizes for vaccine coverage > 25% ) . This effect arises because interrupting transmission renders vaccination the sole source of cross-protective immunity in the population . The incremental impact of increased vaccination coverage is thus greater than in regimes allowing transmission , where infection is an additional source of cross-protective immunity . To additionally explore the validity of these findings under parameter uncertainty , we conduct a multivariate sensitivity analysis as described in the Methods . In particular , we explore the key outputs of this analysis: the relative performance of conventional and universal vaccines , with respect to control of seasonal influenza; impact on the pace of antigenic evolution; and implications for pandemic control . Taking the first of these as an example , if gC is the average seasonal epidemic size under a given vaccination coverage , and gU is the corresponding quantity for a universal vaccine , we calculate the ratio r = gU / gC . As long as this quantity is below 1 , the qualitative finding in Fig 3A holds true . Defining r as the ‘relative efficiency’ in control of seasonal epidemics , we likewise consider relative efficiencies in controlling antigenic evolution , and in pandemic control ( corresponding to each of the panels in Fig 3 ) . Fig 4 plots these relative efficiencies , together with their uncertainty , for different levels of vaccination coverage . In each panel , the region above the dashed line ( i . e . a ratio > 1 ) corresponds to conventional vaccines being more efficient than universal vaccines , and vice versa . Fig 4B and 4C suggest that universal vaccines are robustly more efficient in controlling antigenic evolution and in mitigating pandemic sizes . Notably , however , the uncertainty bounds in Fig 4A straddle the line r = 1 ( shown ‘dashed’ ) , indicating certain parameter combinations under which a universal vaccine could allow greater seasonal epidemics than a conventional vaccine . To identify which parameters are driving this result , taking a vaccination coverage of 15% , Fig 5 shows scatter plots of the relative efficiency r with respect to each of the parameters in the model . Points of interest ( r > 1 ) are shown in red , and are roughly evenly distributed for each of the parameters , with the notable exception of h ( fourth panel ) , where values of r > 1 clearly cluster around a low duration of protection . Motivated by this Figure , holding h constant at its default value , and re-sampling other parameters , yields values of r strictly less than 1 ( see Fig C in S1 Appendix ) . Overall , therefore , in the range of parameter values explored here , universal vaccines appear robustly more efficient in controlling seasonal epidemics , as long as the duration of protection that they provide is sufficiently long .
While a major focus in the development of new influenza vaccines is on their ability to provide individual protection , anticipating the population-level effects of vaccination can also yield useful public health insights . Here , we present a simple model bringing together influenza evolution and epidemiology , and use this model to compare vaccination programmes with two different types of influenza vaccine: current , ‘conventional’ strain-matched vaccines , versus emerging , ‘universal’ vaccines . A primary result from this work is the contrasting evolutionary effects associated with the two types of vaccines . In general , sustained control of transmission reduces the number of immune individuals in the population , and thus dampens selection pressure for new antigenic variants ( Fig 3B ) : universal vaccines , because they are not HA-specific , are able to achieve this state without themselves contributing to HA selection pressure . Thus the pace of antigenic evolution decreases with higher universal vaccine coverage . However , conventional vaccines raise strain-matched immunity: they therefore have the opposite effect to universal vaccines , compounding HA selection pressure and thus tending to accelerate antigenic evolution . In control of seasonal influenza , universal vaccines could also avert more transmission per dose administered than conventional vaccines , with the potential to interrupt transmission even at moderate levels of coverage ( Fig 3A ) . This amplified effect likely arises from the fact that universal vaccination reduces strain-matched immunity in the population while increasing cross-protective immunity and slowing antigenic drift , while conventional HA-specific vaccines do the converse . Overall , therefore , universal vaccination could complement population immunity in a way that is more efficient for controlling transmission , over several seasons , than strain-matched vaccines . Furthermore , while the effects of HA-specific vaccination are limited by antigenic evolution , the effects of universal vaccination are limited by the duration of cross-protective immunity [28] . As long as this duration is long enough to persist across vaccination intervals , the effect of cross-protective vaccination can be maintained on a population level with each passing season . Conventional vaccines provide HA-specific immunity against seasonally circulating strains at the expense of infection-acquired immunity that may otherwise protect against novel antigenic subtypes [48–50] . Thus , as shown in Fig 3C , increased HA-specific vaccination coverage could result in increased pandemic sizes . Indeed , these model findings are consistent with experimental findings in animal challenge studies [50] . By contrast , a universal vaccine , even if transmission-blocking rather than infection-blocking , could reduce pandemic sizes by promoting cross-protective immunity in the population . Similar phenomena have been suggested by Zhang et al via different mechanisms , by which cross-protective immunity limits the opportunities for reassortment , thus limiting the emergence of pandemic-capable viruses ( [28] ) . Taken together , these findings suggest that universal vaccines could be effective in both preventing and mitigating pandemic emergence . While influenza is a readily evolving pathogen , it is evidently not so rapidly evolving as to cause pandemic-scale epidemics every season . Here , we capture this phenomenon by assuming that viral evolution is limited by the available HA diversity in the population ( Fig 1B ) . As for the conserved antigens targeted by universal vaccines , we have ignored the potential for immune escape , assuming in this work that any antigenic change would be too functionally costly for the virus to continue replicating . Nonetheless , the potential for such immune escape cannot be discounted: should it occur it would have far reaching consequences , comparable to pandemic emergence . Additionally , even if conventional vaccines should have negative implications for pandemic control , for their sterilizing immunity they would remain essential in routine immunization to protect specific risk groups such as the immunocompromised and the elderly . Overall then , rather than replacing one vaccination programme with another , it is important to consider universal vaccines as being strategically complementary to conventional , strain-matched vaccines . With recent work highlighting the potential effects of influenza vaccination programs in controlling transmission [25] , our work suggests that—depending on the characteristics of new vaccines including duration of protection and vaccination frequency—the ‘transmission dampening’ role could be one best filled by universal vaccines . An alternative could be a ‘cocktail’ formulation consisting of a combination of strain-specific and vectored , cross-protective immunogens . Such cocktails could continue to protect clinical risk groups such as the elderly , as well as maintaining cross-protective immunity in the population to mitigate pandemic risk . However , their effect on seasonal influenza evolution would depend on the relative strengths of strain-specific and cross-protective protection that they provide: future work could explore the extent to which the cross-protective component of a cocktail vaccine could mitigate the potential ‘evolution-speeding’ effects of its strain-specific component . The present model has several limitations to note . First , it involves a stylized model of influenza evolution: in practice , the antigenic dynamics of influenza arise from a combination of complex processes , spanning the chance emergence of an immune escape variant in an infected host; the transmission of that mutant to other hosts; and its successful establishment in the global population , all in competition with other potential escape variants ( [51 , 52] ) . Each of these stages is stochastic , giving rise to notable irregularities in influenza evolution such as antigenic ‘jumps’ shown by influenza A , every 3–9 years , with important consequences for vaccine selection [41] There is also notable variation in the geographical source of circulating influenza strains each year . [53] Nonetheless , the aim of the present work is not to explain such spatiotemporal variation , but rather to capture the essential , long-term interplay between population immunity and viral evolution . Consequently our current findings for universal vaccines ( particularly , that they could slow antigenic evolution ) are consistent with previous work , which employed a more complex , stochastic framework [29 , 54]: we would expect our current findings for conventional vaccines to be similarly qualitatively robust to stochasticity in viral evolution . Second , the model does not take into account heterogeneities such as age structure [45 , 46] . With school-age children playing an important role in the transmission of influenza ( [55–57] ) , and the elderly being less important for transmission , the effect of a given population coverage of vaccination will depend critically on how it is distributed amongst age groups ( [58] ) . Neglecting such effects , our model may overestimate the impact of a given vaccination coverage , for example , suggesting such low seasonal epidemic sizes at current levels of coverage in the US ( Fig 3A ) . If this bias applies equally for universal as well as for conventional vaccines , it may not be expected to influence our overall results about their relative efficiencies . Moreover , an important area for future work would be the potential impact of age-targeted vaccination programmes for emerging , transmission-controlling vaccines . Several important caveats about immunity also bear mention: first , in the absence of relevant data , we have assumed that vaccine-induced immunity has an efficacy equivalent to its counterpart in natural immunity . Further work could explore the implications of relaxing this assumption . It might be expected that model results would depend to a large extent on whether vaccine-induced immunity would be more or less effective ( or long-lasting ) than its counterpart in natural immunity . ( It is notable , for example , that recombinant technology raises the prospect of focusing immunity on particular antigens to a greater extent than is possible through natural immunity [22 , 59] ) . Second , for simplicity we have neglected the potential for complex interactions such as between antibody-mediated and T-cell mediated immunity , and the potential effect of an individual’s infection history on their vaccine response [60–64] . These complexities are only starting to be explored for influenza , and in future a better understanding of these immunological interactions will allow refined models to explore their implications . Third , we have assumed that universal vaccination does not protect against infection ( i . e . no reduction in susceptibility ) . This being a conservative assumption , we might expect our overall findings to be accentuated by allowing for such additional protection . Conversely , we have assumed that current , strain-matched vaccines elicit no cross-protective immunity . Although this is a helpful caricature for contrasting two different modes of vaccination , conventional vaccines may also elicit some heterosubtypic immunity [65] . In practice , any such protection is unfortunately too weak for current vaccines to protect against novel pandemic strains [32] , a major rationale for universal vaccines [23 , 35]–nonetheless , any broad protection from current vaccines would tend to narrow the gap between conventional and universal vaccines illustrated in Fig 3 . Such caveats notwithstanding , our overall findings are likely to hold true: a vaccine formulation enhancing cross-protective over strain-specific immunity would have qualitatively different population implications from one that does the converse . Overall , a key data need in future is a quantitative comparison of the duration and potency of cross-protection raised by current vaccines , against that offered by emerging vaccine candidates . In summary , emerging vaccine technology , along with increasing interest in understanding the biology of influenza evolution , are offering fresh prospects for the control of influenza . In the context of these and other developments , it is becoming increasingly important to understand the role of the various arms of natural and vaccine-induced immunity in controlling influenza , and in driving viral evolution . By aiming to link these critical host mechanisms to important phenomena on the population level , mathematical models , such as the one presented here , can be valuable in casting light on the potential impact of new and emerging vaccines .
|
Influenza vaccines used today offer good protection , but have limitations: they have to be updated regularly , to remain effective in the face of ongoing virus evolution , and they cannot be used in advance of an influenza pandemic . In this study we considered how such ‘conventional’ vaccines might compare on the population level against new ‘universal’ vaccines currently being developed , that may protect against a broad spectrum of influenza viruses . We developed a mathematical model to capture the interactions between vaccination , influenza transmission , and viral evolution . The model suggests that annual vaccination with universal vaccines could control annual influenza epidemics more efficiently than conventional vaccines . In doing so they could slow viral evolution , rather than promoting it , while maintaining the broadly protective immunity that could mitigate against the emergence of a pandemic . These effects depend sensitively on the duration of protection that universal vaccines can afford , an important quantity to monitor in their development . In future , it is likely that conventional and universal vaccines would be deployed in tandem: we suggest that they could fulfill distinct roles , with universal vaccines being prioritised for managing transmission and evolution , and conventional vaccines being focused on protecting specific risk groups .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"organismal",
"evolution",
"viral",
"vaccines",
"medicine",
"and",
"health",
"sciences",
"influenza",
"immunology",
"cell-mediated",
"immunity",
"microbiology",
"vaccines",
"preventive",
"medicine",
"microbial",
"evolution",
"vaccination",
"and",
"immunization",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"evolutionary",
"immunology",
"viral",
"evolution",
"immunity",
"virology",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"evolutionary",
"biology"
] |
2016
|
Universal or Specific? A Modeling-Based Comparison of Broad-Spectrum Influenza Vaccines against Conventional, Strain-Matched Vaccines
|
Foamy macrophages ( FM ) s harbor lipid bodies that not only assist mycobacterial persistence within the granulomas but also are sites for intracellular signaling and inflammatory mediators which are essential for mycobacterial pathogenesis . However , molecular mechanisms that regulate intracellular lipid accumulation in FMs during mycobacterial infection are not clear . Here , we report for the first time that jumonji domain containing protein ( JMJD ) 3 , a demethylase of the repressive H3K27me3 mark , orchestrates the expression of M . tuberculosis H37Rv- , MDR-JAL2287- , H37Ra- and M . bovis BCG-induced genes essential for FM generation in a TLR2-dependent manner . Further , NOTCH1-responsive RNA-binding protein MUSASHI ( MSI ) , targets a transcriptional repressor of JMJD3 , Msx2-interacting nuclear target protein , to positively regulate infection-induced JMJD3 expression , FM generation and M2 phenotype . Investigations in in vivo murine models further substantiated these observations . Together , our study has attributed novel roles for JMJD3 and its regulators during mycobacterial infection that assist FM generation and fine-tune associated host immunity .
Macrophages recruit to the site/s of mycobacterial infection and differentiate to several cell types including the lipid droplet loaded foamy macrophages ( FMs ) [1] . The characteristic lipid body ( LB ) generation in the FMs provides a suitable survival niche for the mycobacteria in the granuloma [2] . It is known that FMs not only provide the nutrient source for the internalized mycobacteria , but also generate an environment conducive for the bacilli to attain the non-replicating persistence state [3] . More importantly , FMs are the reservoirs for several inflammatory mediators such as arachidonic acid and the enzymes that catalyze the conversion of arachidonic acid to immunomodulatory eicosanoids like cyclooxygenase-2 , lipoxygenases-5 and -15 and LTC4-synthase [1] . Additionally , LBs also regulate lipid metabolism , membrane trafficking and intracellular signaling [4] . Hence , understanding the mechanisms that regulate mycobacterial infection-induced FM generation assumes importance . A complex co-ordinated mechanism of lipid influx , metabolism , storage and mobilization constitute the FM generation [4] . These processes are carried out by a dedicated set of diverse genes . The genes that assist in lipid biosynthesis like acyl-CoA synthetase long-chain family member-1 ( Acsl1 , lipid biosynthesis ) , adipose differentiation-related protein ( Adrp/Plin2 , lipid droplet synthesis ) and sphingolipid activator proteins ( Psap/SapC , its product maintains the turnover of lipids in membranes ) are known to be upregulated during FM formation [1 , 5] . Membrane proteins like fatty acid translocase ( Fat/CD36 ) , macrophage scavenger receptor-1 ( Msr1/CD204 ) and macrophage receptor with collagenous domain ( Marco ) were essential for uptake of low-density lipoproteins into the macrophages , a characteristic evidence for FM generation [3] . The low-density lipoproteins inside the FMs are metabolized to triacylglycerides , phospholipids and cholesterol . The esterified cholesterol either gets sequestered to the LBs or is effluxed via the ATP-binding cassette ( ABC ) transporters , Abca1 and Abcg1 [3] . Hence , deficiency or downregulation of the ABC transporters favour FM generation [6] . Importantly , fine-regulation of the above mentioned genes would orchestrate the FM phenotype and functions during mycobacterial pathogenesis . In this context , regulatory mechanisms governing such pathogen-specific spatio-temporal inflammatory responses would involve reversible , instantaneous but specific action like the ones mediated by epigenetic regulators [7] . Of the various epigenetic mechanisms , histone modifications play vital roles in regulating the gene expression [8] . Interestingly , many histone marks including Histone H3 lysine 27 trimethylation ( H3K27me3 ) have been implicated in inflammation and pathogenesis [9] . It is well established that H3K27me3 brings about the silencing of genes [10] . In general , trimethylation of H3K27 is catalyzed by EZH2 , which associates with SUZ12 , EED and RbAp48 to form the polycomb-repressive complex 2 ( PRC2 ) and jumonji domain containing protein ( JMJD ) 3 is a known H3K27me3 demethylase [10] . Importantly , PRC2 complex is a potent regulator of several signaling pathways like NOTCH1 , WNT and sonic hedgehog signaling [11] which have been reported to be activated during mycobacterial infection to direct the immune responses and determine the cell-fate [12–15] . Additionally , reports have implicated the role for JMJD3 in regulating inflammation and TLR responses [10 , 16 , 17] including generation of M2 phenotype [18] and foamy characteristics of macrophages during atherosclerosis [19] . Of note , M2 macrophages function to exacerbate mycobacterial pathogenesis [15 , 20–22] and FM molecular markers such CD36 , MSR1 , lipoxygenases 5/15 etc constitute M2 macrophages [23 , 24] . In this perspective , the role for H3K27 methylation by PRC2 complex and its demethylase , JMJD3 during mycobacteria-responsive FM generation was explored . Infection of macrophages with M . tuberculosis H37Rv ( represented as H37Rv ) , multi-drug resistant strain MDR-JAL2287 , M . tuberculosis H37Ra ( represented as H37Ra ) or M . bovis BCG ( represented as BCG ) , but not M . smegmatis , displayed JMJD3-dependent LB formation . Supporting this observation , the genes involved in lipid biosynthesis ( Acsl1 , Adrp , Psap ) and uptake ( Fat/CD36 ) and Msr1 ) were significantly upregulated with mycobacterial infection of macrophages in a JMJD3- and TLR2-dependent manner . Deciphering the mechanism of JMJD3 expression , we found an evolutionarily conserved RNA-binding protein MUSASHI ( MSI ) , which was NOTCH1-responsive , to target a transcriptional repressor of JMJD3 , Msx2-interacting nuclear target protein ( MINT/Spen ) and thus assist in FM generation during mycobacterial infection . Immunohistochemistry ( IHC ) and immunofluorescence ( IF ) experiments utilizing H37Rv-infected lungs and in vivo murine BCG-induced granuloma model substantiated these observations . MSI-JMJD3 axis was found to regulate M2 phenotypic responses in the FMs during mycobacterial infection . Thus , the current investigation has identified roles for JMJD3 and associated epigenetic regulators to shape the immune responses during mycobacterial pathogenesis .
FMs are the integral components of granulomas during mycobacterial pathogenesis [2] . However , mechanisms that regulate intracellular lipid accumulation in the FMs during the course of mycobacterial infection require extensive investigation . To begin with , the ability of different mycobacterial species to induce FMs was analyzed . H37Ra- and BCG-infected RAW 264 . 7 macrophages , unlike M . smegmatis , stained positive for LB formation as assessed by Oil Red O ( ORO ) staining and measurement of the extracted ORO ( S1A Fig ) . With reports suggesting a crucial role for TLR2 signaling during FM generation [25] and TLR2 signaling being a major mediator of mycobacterial pathogenesis [26] , analysis of the contribution of TLR2 during mycobacteria-mediated FM formation was undertaken . In this regard , RAW 264 . 7 macrophages expressing the dominant negative form of TLR2 failed to generate FMs on H37Ra and BCG infection ( S1A Fig ) . Additionally , lipid bodies in primary mouse macrophages were labeled with fluorescent dye BODIPY 493/503 to assess the frequency and mean fluorescence intensities ( MFIs ) of FMs . In line with the ORO results , H37Ra and BCG showed increased frequency and MFIs of BODIPY-stained macrophages in WT mice when compared to that in tlr2-null mice ( S1B Fig ) . For the reasons mentioned earlier in the introduction and to understand the mechanisms governing such mycobacteria-TLR2-specific responses , we sought to uncover the role for epigenetic regulation , if any , of the coordinated process of FM generation . Of note , H3K27me3 is a histone mark widely implicated during inflammation and pathogenesis [27] . Hence , H3K27-associated methylase EZH2 and demethylase JMJD3 were analyzed to explore the epigenetic regulation of TLR2-dependend FM induction during mycobacterial infections . Primary macrophages infected with H37Ra or BCG , but not M . smegmatis , were found to display elevated expression of JMJD3 at both RNA and protein levels ( Fig 1A and 1B ) . However , no significant global change in the levels of H3K27me3 and EZH2 were observed ( Fig 1A ) . Interestingly , peritoneal macrophages infected with virulent strains of M . tuberculosis like H37Rv and MDR-JAL2287 induced a robust expression of JMJD3 ( Fig 1C and 1D ) . Like avirulent strain of mycobacteria , H37Rv and MDR-JAL2287 showed increased frequency and MFIs of BODIPY-stained macrophages ( Fig 1E and 1F ) and ORO staining ( S1C Fig ) , indicative of significant FM generation . Role for TLR2 in mediating the mycobacteria-induced JMJD3 expression was verified in macrophages obtained from tlr2-null mice ( Fig 1G and 1H ) . To establish the physiological role for JMJD3 during LB generation , overexpression and knockdown experiments of JMJD3 were performed . While RAW 264 . 7 macrophages overexpressing JMJD3 showed significant induction of the FMs ( Fig 1I and S1D and S1E Fig ) , macrophages depleted of JMJD3 using specific siRNAs displayed compromised ability to generate FMs on mycobacterial infection of H37Rv and BCG ( Fig 1J and S1F and S1G Fig ) . As mentioned earlier , FMs house several inflammatory mediators that contribute to the M2 phenotypic and functional properties of FMs in different disease contexts [23 , 24] . We thus assessed the role for JMJD3 in orchestrating the immune responses exhibited by FMs . Interestingly , while JMJD3 was found to negatively regulate few of the mycobacteria-responsive M1 markers of macrophages viz . , Il12 , Il1b and Cxcl2 ( S2A Fig ) , the expression of M2 markers like Arg1 , Mrc1 , Il10 , Tgfb , Ccl17 and Ccl2 on infection were JMJD3-dependent ( Fig 1K ) . Several genes co-ordinate the lipid biosynthesis , uptake and accumulation processes during FM formation [5] . Expression analysis of few of the genes suggested that H37Ra and BCG , but not M . smegmatis , induced the positive regulators ( Acsl1 , Adrp , Psap , Fat , Msr1 and Marco ) and downregulated or did not modify the negative regulators ( Abca1 and Abcg1 ) of FM generation ( S2B Fig and Fig 2A ) . In accordance with the previous results utilizing virulent mycobacterial strains , H37Rv and MDR-JAL2287 expressed the FM genes ( Fig 2B ) . Further , mycobacteria-responsive expression of Acsl1 , Adrp , Psap , Fat and Msr1 was found to be TLR2-dependent ( Fig 2C ) . Since induced expression of JMJD3 was found essential to render the FM phenotype during mycobacterial infection ( Fig 1J ) , contribution of JMJD3 in regulating the identified set of genes was assessed . Corroborating the previous observation , RAW 264 . 7 macrophages expressing Jmjd3-specific siRNAs failed to express Acsl1 , Adrp , Psap and Fat on BCG infection ( Fig 2D and 2E ) . To further validate the epigenetic regulation of the identified genes , ChIP experiments were performed . BCG infection of primary macrophages showed TLR2-dependency for decreased H3K27me3 methylations on the promoters of Acsl1 , Adrp , Psap and Fat genes marking the active transcription ( Fig 2F ) . Importantly , corresponding recruitment of JMJD3 to the identified promoters was found . While BCG infection of wild-type macrophage showed significant recruitment of JMJD3 to the promoters of Acsl1 , Adrp , Psap and Fat genes , macrophages from tlr2-null mice did not exhibit similar results ( Fig 2F ) . Possibility of direct regulation of the M2 genes by JMJD3 was ruled out as ChIP results suggested that mycobacterial infection does not modulate H3K27me3 modification or recruitment of JMJD3 to the promoters of M2 genes in macrophages ( S2C Fig ) . Additionally , BODIPY and transcript analysis in macrophages transfected with specific siRNAs to Acsl1 , Adrp , Psap and Fat genes underscored the role for these genes in regulating FM generation ( Fig 2G and S2D Fig ) and M2 gene expression ( S2E Fig ) . Together , this suggests that mycobacteria-induced TLR2 signaling directs the JMJD3-dependent expression of genes required for LB formation , FM generation and concomitant M2 phenotypic responses . To bring the in vivo relevance of the identified mechanism of FM generation , we utilized two murine models . Lungs were isolated from WT and tlr2-null mice after aerosol infection with H37Rv . Characteristic lesions were observed in larger numbers on the pleura of infected WT mice as compared to that in infected tlr2-null mice ( S3A Fig and Fig 3A ) . Analysis of lung tissue sections revealed characteristic granulomas with epithelioid cells and lymphocytes in the H37Rv-infected mice . However , no necrosis was observed . While 5–8 such granulomas were observed in the lungs of the WT mice , tlr2-KO mice were found to have 2–3 granulomas in their lungs . Importantly , the granuloma score was significantly reduced from 32 . 5 in the WT to 17 . 5 in the tlr2-KO mice ( Fig 3B and 3C ) , indicating the TLR2 dependency of mycobacteria-induced granulomas . Further , BODIPY analysis of the lungs by IF ( Fig 3D ) and FACS ( Fig 3E ) and ORO staining ( S3B Fig ) validated the TLR2-dependent FM generation in mice during H37Rv infection . In an alternate study , a previously well known in vivo murine granuloma model [5 , 28] was established with BCG infection . The excised granulomas were analyzed for the characteristic hallmarks of a granuloma such as cellular architecture , peripheral accumulation of the lymphocytes and different classes of macrophages constituting the center ( S3C Fig ) . After authenticating the obtained granulomas , ORO staining of the sections were performed . BCG-induced granulomas from wild-type mice displayed increased occurrence of FMs in the tissues when compared to that from tlr2-null mice ( S3D Fig ) . Importantly , results from IHC ( Fig 3F ) and IF ( S3E Fig ) experiments suggested that while H37Rv- and BCG-induced granulomas from wild-type mice express elevated levels of JMJD3 , ADRP and CD36 , expression of these genes in granulomas from tlr2-null mice was significantly abrogated . After establishing the downstream effector functions of JMJD3 , the possible mechanisms by which mycobacteria induces JMJD3 expression were explored . In this context , silencing mediator for RARs and thyroid receptors-extended ( SMRTe/NCoR2 ) is a known repressor of JMJD3 expression in neuronal cells [29] . Importantly , SMRTe complexes with MINT/Spen , HDAC1 and HDAC2 to bring about the repression of the target genes [30] . Hence , the role for SMRTe and MINT/Spen to regulate JMJD3 expression was assessed . As shown in Fig 4A and 4B , while no significant change in the expression of SMRTe was observed , expression of MINT protein was significantly downregulated on H37Rv , H37Ra and BCG infection in primary macrophages . This observation corresponds to the induced expression of JMJD3 in these conditions ( Fig 1A–1D ) . However , levels of Spen transcripts did not alter with the infection ( Fig 4C and 4D ) . Interestingly , we found that inhibition of MINT by specific siRNA results in significant increase in JMJD3 expression , even in the absence of infection ( Fig 4E and 4F ) . To further establish the role for MINT in negatively regulating JMJD3 responses , MINT was overexpressed in RAW 264 . 7 macrophages . Ectopic expression of MINT not only suppressed the ability of mycobacteria ( both BCG and H37Rv ) to induce JMJD3 and the downstream genes responsible for FM generation ( Fig 4G , 4H and 4I ) , but also compromised the ability of BCG to form FMs ( Fig 4J and 4K ) . Thus , mycobacteria subdue the expression of MINT to elevate JMJD3 expression and mediate FM formation . Interestingly , as shown , mycobacteria-mediated inhibition of MINT/Spen was observed at the protein but not at transcript levels ( Fig 4A–4D ) . This underscores a regulation-mediated by post-transcriptional modifications . One such known regulatory mechanism is exhibited by a RNA binding protein , MSI [31] . MSI isoforms MSI1 and MSI2 , bind to the 3’UTR of the target mRNA to block its translation [32] . In the current context , Spen 3’UTR was analyzed for the binding site of MSI , ( G/A ) UnAGU ( n = 2–3 ) [32] . Importantly , a binding site ATTAGT spanning the 332–337 residues of the Spen 3’UTR was identified ( Fig 5A ) . Thus , it was hypothesized that mycobacteria may regulate MINT via MSI activity . H37Rv , MDR-JAL2287 , H37Ra and BCG , but not M . smegmatis infection of primary macrophages was found to exhibit elevated expression of MSI1 and MSI2 at both RNA and protein levels in a TLR2-dependent manner ( Fig 5B–5F ) . Substantiating this observation , significant expression of MSI was found in the infected lungs as well as granuloma sections ( Fig 5G and 5H ) . Further , BODIPY and ORO staining of RAW 264 . 7 macrophages expressing MSI overexpression construct ( Fig 5I and S4A and S4B Fig ) or MSI dominant negative ( S4C and S4D Fig ) and BODIPY analysis of primary macrophages transfected Msi-specific siRNA ( Fig 5J and S4E Fig ) suggested that MSI expression was crucial to mediate mycobacteria-induced FM generation . Supporting this observation , the genes regulating FM formation , Jmjd3 , Acsl1 , Adrp , Psap and Fat were positively regulated by MSI ( Fig 6A , 6B and 6C ) . Interestingly , expression of MINT , a putative target of MSI , was not only suppressed in RAW 264 . 7 macrophages overexpressing MSI , but also rescued in macrophages expressing MSI dominant negative despite the infection ( Fig 6A and 6C ) . To further validate the direct interaction of MINT with MSI , RNA IP experiments were performed . Numb is known target of MSI and was used as a positive control in the experiment [31] . Importantly , the MSI immunoprecipitates from RAW 264 . 7 macrophages infected with BCG or from RAW 264 . 7 macrophages overexpressing MSI showed significant enrichment of MSI binding region from 3’UTR of Spen and Numb ( Fig 6D ) . However , BCG-induced enrichment of MSI binding region from Spen and Numb 3’UTR was severely abolished in macrophages expressing MSI dominant negative ( Fig 6D ) . Together , these results establish that MINT is bonafide target of MSI . We further assessed the contribution of MSI in regulating the immune responses displayed by FMs . In accordance with JMJD3 , MSI negatively regulated M1 markers like Il12 , Il1b and Cxcl2 ( Fig 6E , left panel ) and was necessary for mycobacterial infection-induced expression of M2 markers like Arg1 , Mrc1 , Il10 , Tgfb , Ccl17 and Ccl2 ( Fig 6E , right panel ) . To establish the signaling link between TLR2 and MSI , a screen for various signaling pathways , previously known to be activated during mycobacterial infection [12–14 , 33] was performed . Macrophages treated with specific inhibitors of NOTCH1 activation ( GSI ) , PI3K ( LY294002 ) , mTOR ( Rapamycin ) , NF-κB ( BAY 11–7085 ) , SHH signaling ( Cyclopamine , Betulinic Acid ) and WNT signaling ( IWP-2 , FH535 ) suggested a role for well-established NOTCH1-PI3K-mTOR-NF-κB pathway in regulating BCG-induced MSI expression ( Fig 7A ) . Following inhibition of NOTCH1-PI3K-mTOR-NF-κB pathway with specific pharmacological inhibitors , the MFI and number of BODIPY-stained lipid bodies in H37Rv-infected primary macrophages were significantly reduced ( Fig 7B and 7C ) . It has been well characterized that activation of NOTCH1 signaling is marked by cleavage of the intracellular domain of the NOTCH1 receptor to form NICD that transduces the downstream signaling . BCG infection-induced NOTCH1 signals via PI3K-mTOR-NF-κB cascade in macrophages in a TLR2-dependent manner ( Fig 7D , left panel and [12 , 33] ) . Further , BCG-infected macrophages expressing Notch1-specific siRNAs failed to activate the downstream PI3K-mTOR-NF-κB pathway ( Fig 7D , right panel ) . In line with Fig 7A , BCG-regulated expression of MSI , MINT , JMJD3 , genes associated with FM generation ( Fig 7E–7H ) and FM phenotype as assessed by BODIPY staining ( Fig 7I ) was found to be significantly reduced in macrophages expressing Notch1-specific siRNAs . These results underscore the NOTCH1 signaling functions that mediate mycobacteria-induced FM generation . Corroborating these results , RAW 264 . 7 macrophages stably expressing NICD exhibited the activation of PI3K-mTOR-NF-κB pathway ( S5A Fig ) , induced comparable level of MSI , JMJD3 , genes associated with FM generation and inhibited MINT expression ( Fig 8A and S5B Fig ) . Additionally , similar to BCG-infected macrophages , NICD-expressing RAW 264 . 7 macrophages exhibited significant ORO staining ( S5C Fig ) . Finally , inhibition of PI3K-mTOR-NF-κB pathway in RAW 264 . 7 macrophages stably expressing NICD significantly hampered their ability to induce the expression of MSI , JMJD3 , genes associated with FM generation and FM phenotype ( Fig 8B , 8C and 8D ) . Together , these results suggest the role for NOTCH1 signaling during LB generation and FM formation during mycobacterial infection .
In patients with M . tuberculosis infection , the bacilli were chiefly found to reside in the lipid-rich environment of FMs [34] . Further , both in vitro and in vivo studies in murine [28] and human granuloma models [34] have underscored the importance of the macrophage-derived FMs in regulating the course of mycobacterial pathogenesis . In accordance with previous observations [23 , 34] , in the current study , H37Rv , MDR-JAL2287 , H37Ra and BCG , but not M . smegmatis , a saprophyte , were found to stimulate FM generation . In vivo generation of FMs in the granuloma was also observed . However , though several host [5 , 23 , 35] and bacterial [34 , 35] components have been identified to regulate the infection-induced FM generation , no studies have attempted to unveil the epigenetic regulation that mediate LB and FM formation . We identified role for a histone demethylase , JMJD3 , in orchestrating the mycobacterial infection-induced FM generation . Functions of an inducible demethylase , JMJD3 has been implicated in case of several viral infections [36] , bacterial effector functions [16 , 37] , inflammation [38] and M2 polarization [18] . Interestingly , pathogenic mycobacterial infection is well characterized for the generation of alternatively activated M2 macrophages , which could aid the bacterial survival and immune evasion [15 , 20–22] . Importantly , many FM characteristic proteins like CD36 , MSR1 , lipoxygenases 5/15 etc are hallmark markers of M2 macrophages indicating a close link between M2 macrophages and FMs [23 , 24] . In this context , the role for JMJD3 in mycobacteria-induced FMs was explored . JMJD3 was indeed , for the first time , found to coordinate the H37Rv- , MDR-JAL2287- , H37Ra- or BCG-induced FM formation and consequent M2 phenotype of the FMs . A recent report indicated an important role for JMJD3 during serum amyloid A-enhancement of oxidized LDL-induced macrophage FM generation [19]; however , there was no mechanism established in the study . Of note , the genes that facilitate FM generation like Acsl1 , Adrp , Psap and Fat exhibited elevated expression on infection with mycobacteria . Though the expression of Msr1 and Marco was induced with infection , no significant changes were observed in tlr2-null macrophages or in the absence of Jmjd3 , H3K27me3 demethylase . Hence , these genes were not pursued further . Interestingly , both virulent mycobacterial strains like H37Rv , MDR-JAL2287 and avirulent strains like H37Ra , BCG induced the robust expression of FM genes , but not M . smegmatis . We also found a robust expression of JMJD3-responsive FM genes on H37Rv or MDR-JAL2287 infection when compared to that with H37Ra and BCG . This could be attributed to the virulence characteristics of H37Rv and MDR-JAL2287 . Further investigation on these aspects is underway . Also , the efflux coordinators , Abca1 and Abcg1 were downregulated or remained unchanged with H37Ra and BCG infection but were significantly induced during M . smegmatis infection . Though it needs further examination , increased ABCA1 and ABCG1 transporters in response to M . smegmatis could facilitate the efficient efflux of cholesterol from the infected macrophages and hence contribute to the reduced frequency of FMs during M . smegmatis infection . Together , in case of H37Rv , MDR-JAL2287 , H37Ra and BCG , results indicate a concerted modulation of the lipid biosynthesis and uptake genes by a master regulator , JMJD3 . Further , mycobacteria-induced post-transcriptional regulator , MSI was found to target a negative regulator of JMJD3 , MINT to facilitate the FM generation . In accordance to this observation , MINT was found to inhibit adipogenic differenciation [39] and inhibition of MINT could induce adipogenesis [40] . Likewise , we found suppression of MINT during pathogenic mycobacterial infection is requisite for JMJD3 expression and LB generation . Apart from a study that suggests MSI expression on infection with Helicobacter pylori that induces stemness and gastric cancer [41] , no available reports have implicated the role for MSI during infection or inflammation . The functions of MSI are usually attributed to regulate cancer and development [32 , 42] . Hence , a novel role for MSI during mycobacteria-responsive FM formation was elucidated in the current study . Importantly , while few M1 markers like Il12 , Il1b and Cxcl5 were negatively regulated by MSI-JMJD3 axis , characteristic M2 markers were directed by this pathway ( in line with [18 , 43] ) . The MSI-JMJD3-dependent M2 genes IL-10 and TGF-β including PGE2 which constitute FMs can regulate T cell responses including the generation and expansion of regulatory T cells [44 , 45] . Thus , the identified pathway could largely contribute to the evasive responses during mycobacterial infection and suppression of such pathways during infection could confer stronger immunity . Infection-induced NOTCH1-PI3K-mTOR-NF-κB signaling was found to mediate MSI expression . While mycobacterial infection was previously shown to induce TLR2-NOTCH1-PI3K-mTOR-NF-κB pathway to regulate immune responses [12 , 33] , different cancer studies have implicated NOTCH3-dependent MSI1 expression [46] and MSI1-dependent NOTCH activation [47] . Further , early activation of NOTCH1 signaling on mycobacterial infection [12 , 33] is in line with the current observation of NOTCH1 signaling inducing MSI at early time points of infection . Together , TLR2-NOTCH1-dependent MSI induction was found to regulate the expression of a demethylase , JMJD3 by suppressing MINT . JMJD3 orchestrated the expression of the genes , Acsl1 , Adrp , Psap and Fat to direct the formation of LBs and FMs ( Fig 8E ) . Confounding studies attribute both anti-bacterial and pro-bacterial functions of TLR2 during mycobacterial infection [15 , 48–50] . Importantly , tlr2-null mice were previously reported to exhibit exaggerated immune responses to high dose ( 500 CFUs , as in the current study ) mycobacterial infection but displayed dispersed granuloma , reduced bacterial clearance and succumbed to infection [51] . Elevated inflammatory phenotype was entailed to TLR2-dependent recruitment of Foxp3+ regulatory T cells to lungs that was compromised in tlr2-null mice [52] . In the current investigation , however , we found a pro-mycobacterial role for TLR2 , as a requirement to establish the FMs in granuloma . Due to the fact that TLR2 effectuates multiple immune responses during mycobacterial infection , the exact contribution of TLR2-dependent JMJD3 in vivo in terms of regulating mycobacteria-induced immune responses needs to be assessed in jmjd3-null mice . Since loss of jmjd3 causes perinatal lethality in mice [53] , this study needs alternative strategies . Our work has underscored the novel functions of epigenetic regulators like JMJD3 during mycobacteria-induced generation of a survival niche like FMs in granulomas and fine-tuning the concomitant immune responses . These regulators could be potential candidates for host-directed therapies against mycobacterial infection .
Primary macrophages were isolated from peritoneal exudates of C57BL/6J , C3H/HeJ and tlr2-KO mice that were purchased from The Jackson Laboratory and maintained in the Central Animal Facility , Indian Institute of Science ( IISc ) . Briefly , mice were intraperitoneally injected with 1 ml of 8% Brewer thioglycollate . After 4 d of injection , mice were sacrificed and peritoneal cells were harvested by lavage from peritoneal cavity with ice-cold PBS . The cells were cultured in DMEM ( Gibco-Invitrogen/Thermo Fisher Scientific ) containing 10% FBS ( Gibco-Invitrogen/Thermo Fisher Scientific ) for 6 to 8 h and adherent cells were used as peritoneal macrophages . Murine RAW 264 . 7 macrophage-like cells obtained from the National Center for Cell Sciences , Pune , India . M . tuberculosis H37Rv and MDR-JAL2287 were kind research gifts from Dr . Kanury V . S . Rao , ICGEB , India . All studies involving virulent mycobacterial strains were carried out at the BSL-3 facility at Centre for Infectious Disease Research ( CIDR ) , IISc . M . bovis BCG Pasteur 1173P2 was obtained from Pasteur Institute , Paris , France; M . tuberculosis H37Ra and M . smegmatis were kind research gifts from Dr . P . Ajitkumar , IISc , India . Bacteria were grown to mid-log phase and used at 10 multiplicity of infection ( MOI ) in all the experiments unless mentioned otherwise . All studies involving mice and virulent mycobacterial strains were carried out after the approval from the Institutional Ethics Committee for animal experimentation as well as from Institutional Biosafety Committee . The animal care and use protocol adhered were approved by national guidelines of the Committee for the Purpose of Control and Supervision of Experiments on Animals ( CPCSEA ) , Government of India . General laboratory chemicals were obtained from Sigma-Aldrich or Merck Millipore . Anti-β-ACTIN and anti-HA antibodies were purchased from Sigma-Aldrich . Anti-H3K27me3 , anti-EZH2 , anti-JMJD3 , anti-MUSASHI ( MSI ) , anti-NUMB , anti-NOTCH1 , anti-Cleaved Notch1 ( Val1744 ) ( NICD ) , anti-Tyr485 p85/ Tyr199 p55 phospho-PI3K , anti-Ser2448 phospho-mTOR , anti-Thr389 phospho-p70S6K and anti-Ser536 phospho-NF-κB p65 were purchased from Cell Signaling Technology . Anti-ADRP , anti-CD36 , anti-ABCA1 , anti-SMRTe and anti-MINT ( SPEN ) were purchased from Santa Cruz Biotechnology , Inc . HRP conjugated anti-rabbit IgG and anti-mouse IgG and anti-rabbit DyLight 488 were obtained from Jackson ImmunoResearch . Fluorescein isothiocyanate ( FITC ) -conjugated monoclonal antibodies ( mAbs ) to mouse MHC class II , phycoerythrin ( PE ) -conjugated mAbs to mouse F4/80 were from BD Biosciences . Anti-mouse CD19-APC and CD3-FITC were from Imgenex . Ziehl-Neelsen ( ZN ) staining Kit was purchased from HiMedia and 4′ , 6-Diamidino-2-phenylindole dihydrochloride ( DAPI ) was from Sigma-Aldrich . BODIPY 493/503 ( 4 , 4-Difluoro-1 , 3 , 5 , 7 , 8-Pentamethyl-4-Bora-3a , 4a-Diaza-s-Indacene ) and HCS LipidTOX Red neutral lipid stain was from Molecular Probes ( Invitrogen/Thermo Fisher Scientific ) . In all experiments , cells were treated with the given inhibitor ( from Calbiochem ) for 1 h before experimental treatments at following concentrations: GSI ( 10 μM ) , LY294002 ( 50 μM ) , Rapamycin ( 100 nM ) , BAY 11–7085 ( 10 μM ) , Cyclopamine ( 10 μM ) , Betulinic Acid ( 10 μM ) , IWP-2 ( 5 μM ) , FH535 β-CATENIN and TCF inhibtor ( 15 μM ) . DMSO at 0 . 1% concentration was used as the vehicle control . In all experiments involving pharmacological reagents , a tested concentration was used after careful titration experiments assessing the viability of the macrophages using the MTT ( 3- ( 4 , 5-Dimethylthiazol-2-yl ) -2 , 5-diphenyltetrazolium bromide ) assay . RAW 264 . 7 macrophages stably expressing NOTCH intracellular domain ( NICD ) were generated as described previously [12] . Briefly , RAW 264 . 7 cells were transfected with pCMV-NICD cDNA construct or pCMV alone using Lipofectamine 2000 ( Invitrogen/Thermo Fisher Scientific ) . Cells were selected in G418 sulfate ( 400 μg/ml ) and screened for NICD expression by immunoblotting as well as assessed for the expression of Hes1 mRNA , a transcriptional target for NOTCH1 by quantitative real-time RT-PCR . Transiently transfection of RAW 264 . 7 macrophages with 5 μg of dominant negative mutant forms of TLR2 , MSI or overexpression constructs of JMJD3 , MINT and MSI was performed using low m . w . polyethylenimine ( Sigma-Aldrich ) . In case of experiments involving siRNA , RAW 264 . 7 macrophage cells were transfected with 100 nM siRNA . Jmjd3 , Acsl1 , Adrp , Fat , Psap , Spen , Msi , Notch1 , non-targeting siRNA and siGLO Lamin A/C were obtained from Dharmacon as siGENOME SMARTpool reagents , which contain a pool of four different double-stranded RNA oligonucleotides . Transfection efficiency was found to be 70–80% in all the experiments as determined by counting the number of siGLO Lamin A/C positive cells in a microscopic field using fluorescent microscope . Further , 48 h post-transfection ( for experiments with RAW 264 . 7 cells ) or 24–36 h post-transfection ( for experiments with peritoneal macrophages ) , the cells were treated or infected as indicated and processed for analysis . Macrophages were treated or infected as indicated and total RNA from macrophages was isolated by TRI reagent ( Sigma-Aldrich ) . 2 μg of total RNA was converted into cDNA using First strand cDNA synthesis kit ( Applied Biological Materials Inc . ) . Quantitative real-time RT-PCR was performed using SYBR Green PCR mixture ( KAPA Biosystems ) for quantification of the target gene expression . All the experiments were repeated at least three times independently to ensure the reproducibility of the results . Gapdh was used as internal control . The primers used for quantitative real-time RT-PCR amplification are summarized in S1 Table . Immunoblotting was performed as previously mentioned elsewhere [15] . Infected or treated macrophages were lysed in RIPA buffer constituting 50 mM Tris-HCl ( pH 7 . 4 ) , 1% NP-40 , 0 . 25% Sodium deoxycholate , 150 mM NaCl , 1 mM EDTA , 1 mM PMSF , 1 μg/ml of each aprotinin , leupeptin , pepstatin , 1 mM Na3VO4 and 1 mM NaF . Equal amount of protein from each cell lysate was resolved on a 12% SDS-polyacrylamide gel and transferred to polyvinylidene difluoride membranes ( PVDF ) ( Millipore ) by the semi-dry transfer ( Bio-Rad ) method . Nonspecific binding was blocked with 5% nonfat dry milk powder in TBST [20 mM Tris-HCl ( pH 7 . 4 ) , 137 mM NaCl , and 0 . 1% Tween 20] for 60 min . The blots were incubated overnight at 4°C with primary antibody followed by incubation with anti-rabbit-HRP or anti-mouse-HRP secondary antibody in 5% BSA for 2 h . After washing in TBST , the immunoblots were developed with enhanced chemiluminescence detection system ( Perkin Elmer ) as per manufacturer’s instructions . β-ACTIN was used as loading control . ChIP assays were carried out using a protocol provided by Upstate Biotechnology and Sigma-Adrich with certain modifications . Briefly , macrophages were fixed with 3 . 6% formaldehyde for 15 min at room temperature followed by inactivation of formaldehyde with addition of 125 mM glycine . Nuclei were lysed in 0 . 1% SDS lysis buffer [50 mM Tris-HCl ( pH 8 . 0 ) , 200 mM NaCl , 10 mM HEPES ( pH 6 . 5 ) , 0 . 1% SDS , 10 mM EDTA , 0 . 5 mM EGTA , 1 mM PMSF , 1 μg/ml of each aprotinin , leupeptin , pepstatin , 1 mM Na3VO4 and 1 mM NaF] . Chromatin was sheared using Bioruptor Plus ( Diagenode ) at high power for 40 rounds of 30 sec pulse ON/45 sec OFF . Chromatin extracts containing DNA fragments with an average size of 500 bp were immunoprecipitated using JMJD3- or H3K27me3-specific antibodies or rabbit preimmune sera complexed with Protein A agarose beads ( Bangalore Genei ) . Immunoprecipitated complexes were sequentially washed [Wash Buffer A: 50 mM Tris-HCl ( pH 8 . 0 ) , 500 mM NaCl , 1 mM EDTA , 1% Triton X-100 , 0 . 1% Sodium deoxycholate , 0 . 1% SDS and protease/phosphatase inhibitors; Wash Buffer B: 50 mM Tris-HCl ( pH 8 . 0 ) , 1 mM EDTA , 250 mM LiCl , 0 . 5% NP-40 , 0 . 5% Sodium deoxycholate and protease/phosphatase inhibitors; TE: 10 mM Tris-HCl ( pH 8 . 0 ) , 1 mM EDTA] and eluted in elution buffer [1% SDS , 0 . 1 M NaHCO3] . After treating the eluted samples with RNase A and Proteinase K , DNA was precipitated using phenol-chloroform-ethanol method . Purified DNA was analyzed by quantitative real time RT-PCR . All values in the test samples were normalized to amplification of the specific gene in Input and IgG pull down and represented as fold change in modification or enrichment . All ChIP experiments were repeated at least three times and the primers utilized are listed in S1 Table . H37Rv was grown in Middlesbrook 7H9 medium ( Difco ) containing 0 . 2% glycerol , 0 . 05% Tween 80 and 10% ADC . Cultures were grown at 37°C to log phase . Culture was washed with PBS and passed 10 times each through 26- , 29- and 31-gauge needles to make single cell suspensions of H37Rv . C57BL/6 and tlr2 null mice ( n = 6 , each group ) were infected with 500 CFUs of H37Rv via aerosol using an aerosol chamber ( Wisconsin-Madison ) . The pulmonary infection dose was confirmed by plating the homogenized lung tissue on Middlebrook 7H10 ( Difco ) agar plates . Post 8 weeks of aerosol infection , mice were sacrificed; lungs were collected , fixed overnight with 3 . 6% formaldehyde ( Sigma-Aldrich ) and processed for cryotomy or microtomy . Sections of fixed mice lungs were stained with hematoxylin and eosin to assess the pathology . Granulomas features were characterized and assigned different scores: with necrosis ( Score = 5 ) , without necrosis ( Score = 2 . 5 ) , with fibrosis ( Score = 1 ) . Total granuloma scores were calculated by multiplying the characterized feature score with the number of granuloma in each lung . C57BL/6 and tlr2 null mice ( n = 7 , each group ) were used for generating granulomas as previously described [3 , 5 , 28] with certain modifications . BCG ( 107 ) were resuspended in 300 ml of ice-cold growth-factor reduced matrigel ( Sigma-Aldrich ) . The mixture was injected sub-dermally into the skin fold at the scruff of the neck . Both sets of mice received either matrigel alone or BCG mixed matrigel . Granulomas were excised at 7 day post-inoculation and processed for cryotomy or microtomy . The excised granuloma was rapidly frozen in liquid nitrogen in the optimal cutting temperature ( OCT ) media ( Jung , Leica ) . Cryosections of 10–15 μm were taken in Leica CM 1510 S or Leica CM 3050 S cryostat ( both Leica ) with the tissue embedded in OCT onto the glass slides and stored at -80°C . Cells/cryosection tissues were fixed in 3 . 6% formaldehyde for 15 min . After PBS wash , the cells/tissues were rinsed in 60% isopropanol for 10 min and stained with 0 . 5% ORO for 15 min . After cleaning the cells/tissues with 60% isopropanol , they were either counterstained with haematoxylin and viewed under 100X or 20X in light microscope or washed in 100% isopropanol for colorimetric analysis at 510 nm . Lipid body staining was performed using a protocol provided by the manufacturer with certain modifications . The cells/cryosection tissues were fixed with 3 . 6% formaldehyde for 45 min . After 3 washes with PBS , the cells/cryosection tissues were stained with 10 μg/ml BODIPY 493/503 or 1X HCS LipidTOX Red neutral lipid stain for 30 min in dark . After 3 washes with PBS , the cells/sections were stained with DAPI and mounted on a slide with glycerol as the medium . Confocal images ( Z-stacks ) were taken on Zeiss LSM 710 Meta confocal laser scanning microscope ( Carl Zeiss AG ) using a plan-Apochromat 63X/1 . 4 Oil DIC objective ( Carl Zeiss AG ) and images were analyzed using ZEN 2009 and ImageJ softwares . The lipid bodies were counted in over 250 cells from different fields . Frequency of populations with different number of lipid bodies ( 0–5 in blue , 5–10 in green , >10 in red ) were plotted in terms of percentage . Statistical significance is represented in green and red . Green line represents the significance analysis of populations with “5–10” lipid bodies between indicated treatments and red represents the significance analysis of populations with “>10” lipid bodies . For the MFI analysis , ImageJ was utilized to calculate the maximum intensity projections of the Z-stacks . Using free hand selection tool , cells were selected to measure the area-integrated intensity and mean grey value . The area around the cells without fluorescence was used to calculate the background values . Corrected Total Cell Fluorescence ( CTCF ) was calculated using the following formula: CTCF = Integrated intensity– ( area of selected cell X Mean florescence of background reading ) . Uninfected and H37Rv-infected lungs were finely chopped and digested in 5 ml RPMI with 5% FBS containing 150 U/ml Collagenase IV ( HiMedia ) and 50 U/ml DNaseI ( Thermo Fisher Scientific ) at 37°C in shaking condition for 1 h . The cell suspension was passed through 40 μm cell strainer and pelleted at 1500 rpm for 10 min . The cells were resuspended in RBC lysis buffer . Post lysis , the cells were washed with PBS . Obtained cell suspension was fixed using 3 . 6% formaldehyde solution for 30 min . After 3 washes with PBS , cells were resuspended in PBS containing 2 μg/ml BODIPY 493/503 and incubated for 30 min at room temperature in dark under mild rocking condition . After thorough washes with PBS , cells were analysed by flow cytometry wherein 1 lakh events were recorded for each sample acquired in BD FACSCanto II . The data was analyzed using FACSDiva software ( BD Biosciences ) and WinMDI Version 2 . 8 . In case of MINT validation experiments , RAW 264 . 7 macrophages-transfected with MINT-EGFP were seeded on to coverslips and incubated . The cells were fixed with 3 . 6% formaldehyde for 15 min at room temperature and the coverslips were mounted on a slide with glycerol . For IF of the cryosections , frozen sections were thawed to room temperature and fixed with 3 . 6% formaldehyde . After blocking with 2% BSA containing saponin , the sections were stained for specific antibodies at 4°C overnight . The sections were incubated with DyLight 488-conjugated secondary antibody for 2 h and nuclei stained with DAPI . A coverslip was mounted on the section with glycerol as the medium . Confocal images were taken on Zeiss LSM 710 Meta confocal laser scanning microscope ( Carl Zeiss AG ) using a plan-Apochromat 63X/1 . 4 Oil DIC objective ( Carl Zeiss AG ) and images were analyzed using ZEN 2009 software . Microtome sections ( 4 μm ) were obtained from 3 . 6% formaldehyde-fixed , decalcified , and paraffin-embedded tissues using Leica RM2245 microtome ( Leica ) . These sections were first deparaffinized , subjected to antigen retrieval by boiling in 10 mM citrate buffer ( pH 6 . 0 ) for 10 min , treated with 1% H2O2 for 10 min , and blocked with 5% BSA for 1 h at room temperature . The tissue sections were further incubated with primary antibodies overnight . After incubation with anti-rabbit HRP-conjugated secondary antibody for 90 min , sections were stained with 0 . 05% diaminobenzidine ( Sigma-Aldrich ) in 0 . 03% H2O2 solution and counterstained with hematoxylin , dehydrated and mounted . Stained tissue sections were imaged with Axio Scope . A1 microscope ( Zeiss ) at indicated magnification . All experiments were performed with appropriate isotype-matched control antibodies . Cryosections were fixed , paraffin-embedded sections were deparaffinized and hydrated to distilled water . The sections were stained with hot Carbol fuchsin solution for 5 min . The sectioned were washed in running water and destained with 1% acid alcohol . After washing with running water , the sections were counter-stained with methylene blue for 30 sec . The sections were washed and dried . The images were acquired with Axio Scope . A1 microscope ( Zeiss ) at indicated magnification . Macrophages were lysed in 300 μl of complete polysomal lysis buffer [5 mM MgCl2 , 0 . 1 M KCl , 0 . 5% NP40 , 0 . 01M HEPES pH 7 . 5] . Total cell lysate ( 200 μg ) was diluted to 500 μl using complete polysomal lysis buffer for IP . 50 μl of anti-mouse IgG precleared lysate was used as Input for the experiment and total RNA was isolated as described earlier . Rest of the precleared lysate was incubated with 1 μg anti-mouse IgG or anti-MSI prebound Protein A beads overnight . Further , beads were washed in complete polysomal lysis buffer and 25% of the beads were eluted in 5X Laemmli buffer for immunoblotting with anti-MSI . Remaining beads were eluted in TRI reagent and processed for RNA isolation as described earlier . 500 ng of the RNA was converted into cDNA and quantitative real time RT-PCR was performed to analyze Spen and Numb . The primers used are listed in S1 Table . Gapdh from Input was utilized for normalization . Levels of significance for comparison between samples were determined by the Student’s t-test distribution and one-way ANOVA followed by Tukey’s multiple-comparisons . The data in the graphs are expressed as the mean ± S . E for the values from at least 3 or more independent experiments and P values < 0 . 05 were defined as significant . GraphPad Prism 5 . 0 software ( GraphPad Software ) was used for all the statistical analysis .
|
Foamy macrophages ( FMs ) not only provide a suitable survival niche for the mycobacteria in the granuloma but also are reservoirs for several inflammatory mediators that regulate mycobacterial pathogenesis . Hence , understanding the mechanisms that regulate infection-induced FM generation assumes importance . In this investigation , we present empirical evidence to support the role of host epigenetic mechanisms in generating FMs and thus facilitating mycobacterial persistence in vivo . We show that the signaling pathways that mediate mycobacteria-induced expression of JMJD3 , a demethylase of the facultative repression mark , regulate the genes assisting in FM generation . Importantly , the identified pathway could largely contribute to the evasive responses during mycobacterial infection and suppression of such pathways during infection could confer stronger immunity . Together , these regulators could be potential candidates for host-directed therapies against mycobacterial infection .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"blood",
"cells",
"reverse",
"transcriptase-polymerase",
"chain",
"reaction",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"molecular",
"probe",
"techniques",
"granulomas",
"gene",
"regulation",
"immunology",
"immunoblotting",
"notch",
"signaling",
"molecular",
"biology",
"techniques",
"research",
"and",
"analysis",
"methods",
"small",
"interfering",
"rnas",
"lipids",
"white",
"blood",
"cells",
"artificial",
"gene",
"amplification",
"and",
"extension",
"animal",
"cells",
"gene",
"expression",
"molecular",
"biology",
"biochemistry",
"rna",
"signal",
"transduction",
"cell",
"biology",
"nucleic",
"acids",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"macrophages",
"non-coding",
"rna",
"cell",
"signaling",
"polymerase",
"chain",
"reaction"
] |
2016
|
MUSASHI-Mediated Expression of JMJD3, a H3K27me3 Demethylase, Is Involved in Foamy Macrophage Generation during Mycobacterial Infection
|
The neonatal intestine is a very complex and dynamic organ that must rapidly adapt and remodel in response to a barrage of environmental stimuli during the first few postnatal weeks . Recent studies demonstrate that the zinc finger transcriptional repressor Blimp1/Prdm1 plays an essential role governing postnatal reprogramming of intestinal enterocytes during this period . Functional loss results in global changes in gene expression patterns , particularly in genes associated with metabolic function . Here we engineered a knock-in allele expressing an eGFP-tagged fusion protein under control of the endogenous regulatory elements and performed genome wide ChIP-seq analysis to identify direct Blimp1 targets and further elucidate the function of Blimp1 in intestinal development . Comparison with published human and mouse datasets revealed a highly conserved core set of genes including interferon-inducible promoters . Here we show that the interferon-inducible transcriptional activator Irf1 is constitutively expressed throughout fetal and postnatal intestinal epithelium development . ChIP-seq demonstrates closely overlapping Blimp1 and Irf1 peaks at key components of the MHC class I pathway in fetal enterocytes . The onset of MHC class I expression coincides with down-regulated Blimp1 expression during the suckling to weaning transition . Collectively , these experiments strongly suggest that in addition to regulating the enterocyte metabolic switch , Blimp1 functions as a gatekeeper in opposition to Irf1 to prevent premature activation of the MHC class I pathway in villus epithelium to maintain tolerance in the neonatal intestine .
The zinc finger transcriptional repressor Blimp1 originally cloned as a negative regulator of beta-interferon gene expression [1] is known to control cell fate decisions in the developing embryo and adult organism [2] . Blimp1 acting as a master regulator of plasma cell differentiation directly represses expression of key transcription factor genes such as c-Myc , Id3 , CIITA and PAX5 required for B-lymphocyte function and proliferation [3 , 4] . In the early embryo , Blimp1 silences the default somatic pathway and instructs a discrete subset of epiblast cells exposed to the highest levels of BMP4 signaling at the base of the allantois to become primordial germ cells ( PGC ) [5 , 6] . At later stages , Blimp1 regulates development of the posterior forelimb , caudal pharyngeal arches , secondary heart field , and sensory vibrissae [7] . Blimp1 also plays an essential role in placental morphogenesis , governing terminal differentiation of the invasive spiral artery-associated trophoblast giant cells [8] . Considerable data strongly suggests that Blimp1 transcriptional targets are cell type-specific . For example Blimp1 directly represses c-Myc expression to block proliferation in B cells , macrophages , and sebaceous gland progenitors [4 , 9 , 10] . In contrast however c-Myc is not a key target in activated effector T cells . Rather Blimp1 blocks IL-2 production required for T cell proliferation [11 , 12] . Within the CD4+ T-cell lineage Blimp1 selectively attenuates Th1 subset development by extinguishing expression of IFNγ , Tbx21 and Bcl6 [13] . Nfat5 , Fos , Dusp16 and Prdm1 itself are direct targets in the skin epidermis [14] . Recent ChIP-seq experiments analyzing transfected P19 embryonic carcinoma ( EC ) cells demonstrate that Blimp1 directly represses numerous developmental and somatic regulators [15] . Cooperative binding with AP2γ and Prdm14 dramatically shifts gene expression profiles and initiates the transcriptional programme required for PGC specification [15] . Blimp1 is strongly expressed in the intestinal epithelium throughout fetal development but beginning at birth becomes dramatically down-regulated in the crypt progenitors , corresponding to the adult intestinal stem cell compartment [7 , 16 , 17] . Conditional deletion experiments revealed an essential role in governing postnatal reprogramming of intestinal enterocytes during the suckling to weaning transition [16 , 18] . Transcriptional profiling experiments demonstrate Blimp1 functional loss results in global changes in gene expression patterns . Thus numerous immature enterocyte markers including digestive enzymes required for processing maternal milk were markedly reduced , whereas in contrast several key components of the adult biochemical signature were substantially and prematurely activated [16] . To further investigate Blimp1 functional contributions during this developmental transition and identify its direct targets in immature intestinal enterocytes , we created a knock-in allele engineered to express an enhanced green fluorescent protein ( eGFP ) -tagged fusion protein reactive with the well-characterized ChIP quality anti-GFP monoclonal antibody . Here we demonstrate that the fusion protein faithfully reconstitutes Blimp1-dependent functional activities . Thus homozygous embryos exclusively expressing the eGFP-tagged Blimp1-fusion protein develop normally , and healthy homozygous adults recovered at the predicted Mendelian ratios were indistinguishable from wild type littermates . Moreover , the knock-in allele efficiently rescues Blimp1-dependent plasma cell differentiation . EGFP-tagged Blimp1 was strongly expressed in the developing intestine allowing us to undertake unbiased ChIP-seq analysis . Several candidate target genes strongly up-regulated in conditional loss mutants were identified as direct targets including Cyp4v3 , Slc16a5 and Myo18b . Interestingly , comparisons of our ChIP-seq peaks with those reported for P19 EC cell and human HeLa cell datasets revealed several highly conserved genomic targets , including key components of the MHC class I peptide-loading pathway [19] . SELEX experiments revealed that the Blimp1 consensus-binding motif closely resembles the IRF-E sequence [20] recognized by IRF1 , an activator of β-interferon gene expression [21–23] . Recent studies suggest that competitive BLIMP1 and IRF1 binding regulates expression of IFN-inducible components of the MHC class I peptide loading machinery [24] . Here we performed genome-wide ChIP-seq analysis to define the extent of overlap between Blimp1 and Irf1 occupancy in vivo under physiological conditions in the intestinal epithelium . We found overlapping Blimp1/Irf1 binding sites proximal to the promoters of 24% ( 24 of 99 ) of human/mouse conserved Blimp1 target genes including Psmb8 , Psmb10 , Psme1 , Tapbp and Erap1 . These findings strengthen the idea that Blimp1 occupancy directly antagonizes Irf1-dependent activation of MHC class I antigen presentation . Consistent with this suggestion we demonstrate that Irf1 is constitutively expressed in the fetal intestine and throughout postnatal stages . The onset of MHC class I expression in the developing intestine precisely coincides with down-regulated Blimp1 expression and the appearance of crypt-derived Blimp1-negative adult enterocytes during the suckling to weaning transition . Besides its role in governing the switch to adult metabolic pathways , Blimp1 co-occupancy at these Irf1-target genes promotes neonatal tolerance in the first weeks after birth during early colonization of the intestinal tract by commensal microorganisms .
To enable identification of Blimp1 target genes in diverse embryonic and adult tissues in vivo under physiological conditions , we engineered a novel eGFP-tagged Blimp1 knock-in allele Prdm1BEG by introducing the cDNA expression cassette into the first coding exon ( Fig 1A ) . This strategy , used successfully for construction of a Prdm1 cre LacZ reporter allele , preserves all known regulatory and structural features of the endogenous locus [8] . Consistent with this homozygous embryos exclusively expressing the Blimp1-eGFP fusion protein develop normally . Additionally , healthy weanlings recovered from intercross matings at the predicted Mendelian ratios , were indistinguishable from wild type littermates ( Fig 1B ) and display no signs of disease when housed in a specific pathogen free environment . Western blot experiments demonstrate the eGFP-tagged fusion protein is strongly expressed in LPS-stimulated splenocytes ( Fig 1C ) . The eGFP-fusion protein efficiently translocates to the nucleus ( Fig 1D ) and rescues plasma cell differentiation ( Fig 1E ) . Robust expression in the spongiotrophoblast and invading spiral artery trophoblast cells in E9 . 5 Prdm1BEG/BEG placenta faithfully reconstitutes Blimp1 functional requirements ( Fig 2A ) [8] . Western blot analysis demonstrates robust expression in the embryonic small intestine ( Fig 2B ) . Immunohistochemical staining confirmed intestinal expression is restricted to the villus epithelium ( Fig 2C ) . Small intestine tissue architecture and body weights were indistinguishable from wild type littermates at all stages examined . Strikingly however , in contrast to mice carrying the exon 1A deletion with modestly reduced expression levels [27] , here we found that healthy adult homozygous animals were sterile . Furthermore , as judged by fast red alkaline phosphatase staining , embryos exclusively expressing the knock-in allele have greatly diminished numbers of PGCs and adult testes lack spermatocytes ( Fig 2D ) . These results demonstrate that the knock-in allele lacks the ability to induce the germ cell transcriptional programme . It is of course possible that functionality of the eGFP-tagged fusion protein is selectively compromised in PGCs due to its inability to recruit the entire cohort of epigenetic partners necessary for silencing of the default somatic pathway . On the other hand , Blimp1 functional requirements in the germ cell lineage are known to be exquisitely dose-dependent [5 , 6] . In all likelihood the failure to rescue germ cell defects simply reflects inadequate expression levels . To identify Blimp1 targets in the developing intestine a well-characterized anti-GFP mouse monoclonal antibody [15 , 28] was exploited for ChIP-seq analysis . We identified 2689 Blimp1 binding events in embryonic ( E18 . 5 ) small intestine ( S1 Dataset ) . Blimp1 binding proximal to selected target genes was validated by ChIP-qPCR ( S1 Fig ) . Analysis of peak locations relative to gene annotations revealed a broad distribution of Blimp1 binding throughout the genome with a bias towards regions proximal to TSSs ( Fig 3A and 3B ) . A comparable number of ChIP-seq peaks ( n = 3018 ) were recently documented in transfected P19 embryonic carcinoma cells over-expressing the identical eGFP-tagged Blimp1 fusion protein [15] . De novo motif analysis of sequences underlying all 2689 Blimp1 peaks revealed a highly significant consensus DNA binding motif ( Fig 3C ) that closely resembles the canonical Blimp1 binding motif originally identified in SELEX experiments [20] , as well as that identified via ChIP-on-Chip analysis of a human myeloma cell line [29] . Functional annotation of genes with proximal Blimp1 binding revealed significant enrichment of GO terms associated with metabolic process and transcriptional regulation ( Fig 3D ) . Comparisons with ChIP-seq peaks recently identified in transfected embryonic carcinoma cells [15] revealed a small subset of overlapping peaks corresponding to roughly 7–8% of each dataset ( Fig 3E ) . The majority ( 171 out of 210 ) fall within 5 kb of annotated gene TSSs . Additionally , comparisons with peaks identified in human HeLa S3 cells [30] revealed 114 highly conserved core Blimp1 genomic targets , including 35 genes ( 37 peaks ) universally present within all three datasets ( Fig 3E and S2 Dataset ) . Interestingly , IFN-inducibility is a common feature shared by several of these components of the MHC class 1 antigen processing machinery i . e . Psmb8 , Psmb10 , Psme2 , Tapbp and Erap1 [19 , 31] . Numerous up-regulated genes previous identified in our transcriptional profiling experiments contain predicted Blimp1 binding sites proximal to their promoter regions [16] . To evaluate whether these candidates are direct Blimp1 target genes , we compared the list of significantly altered genes ( P<0 . 05 with Benjamini and Hochberg multiple testing correction ) with the present ChIP-seq Blimp1 binding events . We observed significant enrichment of Blimp1 binding sites proximal to up-regulated genes ( Fig 3F ) . In contrast , there was no significant overlap between genes with proximal Blimp1 binding and those having decreased expression in conditional mutants . Genes with markedly increased expression that show proximal Blimp1 binding ( S3 Dataset ) include Cyp4v2 , Slc15a5 , Myo18b and Il33 ( increased 27 . 7 , 16 . 1 16 . 5 and 4 . 6 fold in mutants , respectively , Fig 3G ) . Notably , as judged by these criteria , Il33 expressed at high levels in epithelial barrier tissues and believed to play a key role in amplifying innate immunity [32 , 33] , is a direct Blimp1 target . Surprisingly Sucrase isomaltase ( Sis ) ( 365 . 2 fold increased in mutants ) predicted in silico to contain multiple Blimp1 binding sites proximal to its promoter [16] lacks occupancy . Similarly , Arginase type 2 ( Arg2 ) ( 174 . 5 fold increased in mutants ) gave no detectable ChIP-seq peak . In vivo in the context of the E18 . 5 embryonic intestine , compacted chromatin near genomic regions surrounding these potential target sites appears to be inaccessible . We observed robust Blimp1 binding proximal to the c-Myc P1 promoter ( Fig 3G ) . However , as previously reported expression is not significantly altered in Prdm1-/- small intestine [16 , 18] . Indeed , the vast majority of binding events detectable in the present ChIP-seq experiments play no obvious role in Blimp1-dependent transcriptional regulation , and rather seem to reflect so-called neutral occupancy [34] . Blimp1 transcripts are significantly up-regulated in conditional null Prdm1-/- small intestine [16] . Similarly in mouse keratinocytes an intronic Blimp1 binding site downstream of exon 3 mediates a negative feedback loop [14] . However here we failed to detect occupancy at the intronic site described by Magnusdottir et al . [14] . Rather we observe robust binding at the alternative distal 1B promoter region ( Fig 3G ) [27] that drives expression in the yolk sac endoderm [27] and embryonic intestine ( S2 Fig ) . This highly conserved ChIP-seq peak was universally present in both mouse and human datasets [15 , 30] . Interestingly the sea urchin Blimp1 orthologue efficiently binds to the conserved Blimp1 consensus motif [24] and a negative autoregulatory loop upstream of its alternative promoter has been implicated in the specification of endomesodermal territories [35] . Collectively these observations suggest that Blimp1 itself is a direct target of repression but mechanistic details may be cell-type specific . It has been known for many years that the Blimp1 consensus binding motif shares a high degree of sequence overlap with IRF-E core sequence containing GAAAGT/C or GAAACT/C [20 , 36] . We also observed here that the IRF1 DNA binding motif was the second most significant match ( P = 2 . 8 x 10−8 ) after Blimp1 to the de novo identified consensus DNA binding motif underlying all Blimp1 ChIP-seq peaks ( S3 Fig ) . Irf1 expression has been previously documented in purified PGCs [37] . Similarly here , as shown in Fig 4A , we observe weak Irf1 staining in migrating PGCs and the pseudo-stratified epithelium of the primitive gut tube as early as E10 . 5 . Western blot analysis demonstrates Irf1 abundantly expressed in the developing intestine at E16 . 5 ( Fig 4B ) is maintained during the time frame when Blimp1 becomes down regulated between postnatal day 7 and 28 [16] . Strong nuclear staining is readily detectable in both the villus epithelium and the crypts at postnatal stages ( Fig 4C ) . Data from the online EurExpress RNA in situ hybridization ( ISH ) database [38] indicates that Irf1 transcripts are present in the developing gut tube as early as E14 . 5 prior to villus formation ( Fig 4D ) . Irf1 is also constitutively expressed in the developing thymus . Transcriptional profiling data ( Small intestine postnatal development , GEO accession number GDS2989 ) [39] demonstrates high levels of Irf1 and Prdm1 , but importantly the closely related family member Irf2 , known to bind the same IRF-E consensus sequence as Irf1 , is only minimally co-expressed in the intestine ( Fig 4E ) [20 , 36 , 40 , 41] . Next to directly identify Irf1 targets in the E18 . 5 small intestine we performed ChIP-seq analysis . We identified 1996 binding events ( S4 Dataset ) broadly distributed throughout the genome displaying a bias towards regions proximal to TSSs ( S4A and S4B Fig ) [42] . Functional annotation revealed significant enrichment for GO terms associated with immune function ( Fig 5A ) and antigen recognition ( Fig 5E ) . Comparison of our Irf1 ChIP seq peak dataset with genomic coordinates previously reported for mouse LPS-treated BMDC [42] revealed a high degree of conservation ( Fig 5B ) . Nearly half ( 48% ) of our Irf1 peaks display a corresponding peak in the previously reported dataset ( 31% overlap ) . The de novo Irf1 DNA binding motif identified here and that described by Garber et al . [42] were indistinguishable ( Fig 5C ) and closely matches those described for IRF1 in primary human monocytes ( P = 3 . 3 x 10−11 ) [43] , Blimp1 ( P = 3 . 4 x 10−6 ) as well as IRF2 ( P = 5 . 1 x 10−6 ) . Roughly 33% IRF1/BLIMP1 overlap was predicted in human myeloma cells [29] , whereas here Irf1 and Blimp1 ChIP-seq peaks show approximately 12% ( n = 331 , S5 Dataset ) overlap . Functional annotation of the 331 overlapping Irf1/Blimp1 ChIP-seq peaks using GREAT revealed significant enrichment of GO terms associated with MHC class I antigen processing ( Fig 5E ) . To evaluate competitive Blimp1 and Irf1 binding we compared the number of reads underlying Irf1 peaks in Blimp1 mutant versus wild type small intestines . Mutants displayed a significant preference for increased Irf1 occupancy at overlapping Blimp1/Irf1 sites ( Fig 5G and S5 Dataset ) . Conversely , overlapping regions were significantly less likely to show reduced Irf1 occupancy in Blimp1 mutants . To directly examine MHC class I expression in the developing small intestine , we performed Western blot experiments . Irf1 is strongly expressed at E16 . 5 and throughout postnatal intestinal development ( Fig 6A ) . In contrast Blimp1 expression is completely lost by P28 , co-incident with the onset of MHC class I expression . Immunohistochemistry similarly demonstrate MHC class I expression , readily detectable in embryonic Peyer’s patches at E18 . 5 , increases dramatically by P28 ( Fig 6B and 6C ) . These findings strongly suggest that Blimp1 repression of IFN-dependent components of the MHC class I peptide-loading pathway plays a key role in restraining premature activation of host immune responses in utero and early postnatal stages . The ability of Blimp1 to mediate gene silencing and re-organize the chromatin landscape at its target genes depends on recruitment of histone-modifying enzymes [2] . Recent work suggests that postnatal maturation of the intestinal epithelium during the suckling to weaning transition is accompanied by global changes in the epigenetic machinery [44] . Decreased Hdac1 and Hdac2 expression was associated with a modest reduction in histone acetylation . However , protein stability may be compromised due to increased levels of digestive enzymatic activities during intestinal maturation . In snap frozen samples directly lysed in SDS sample buffer we observed consistently high levels of Blimp1 epigenetic partners including Lsd-1 , G9a , Hdac1 and Hdac2 throughout the postnatal period ( S5 Fig ) . Thus dramatic changes in transcriptional profiles during postnatal maturation of the intestinal epithelium cannot simply be explained due to developmentally regulated shifts in the composition of co-repressor complexes . Human placenta architecture , the duration of pregnancy , suckling behaviors and time frames of intestinal maturation have been adapted to fit the lifestyles of different mammalian species . In mice , crypt progenitors emerge from the inter-villous epithelium only after birth . Migration into the underlying mesenchyme leads to the formation of mature crypts and crypt-derived Blimp1-negative adult enterocytes gradually repopulate the intestinal epithelium during the suckling to weaning transition [16] . By contrast in humans overt cryptogenesis begins much earlier . Crypt-like structures initially appear in the small intestine in utero at approximately 12 weeks [45] . Considerable evidence suggests the gut epithelium retains its immature status throughout fetal development . Consistent with this human fetal enterocytes lack ARG2 expression [46] . Additionally a human H4 fetal intestinal enterocyte cell line co-expresses PRDM1 and very low levels of ARG2 and SIS [47] . As shown in Fig 7A , immunohistochemistry demonstrate that BLIMP1 is constitutively expressed throughout the early epithelium at the pseudostratified stage and persists in the forming villous structures at 15 weeks of gestation ( Fig 7B ) . In contrast IRF1 expression is undetectable in the gut epithelium prior to villus formation ( Fig 7C ) . Slightly later at 15 weeks we detect IRF1 co-expressed together with BLIMP1 ( Fig 7D ) . In humans BLIMP1 expression in crypt-like structures at late gestation pregnancy is probably required for continued production of immature enterocytes .
Target gene expression controlled by the zinc finger transcriptional repressor Blimp1/Prdm1 has been extensively characterized in the context of B-cell terminal differentiation to become antibody secreting plasma cells , the development of diverse CD4+ T lymphocyte subsets [4 , 48 , 49] as well as cell fate decisions governing effector versus regulatory CD8+ T cell homeostasis [12 , 50–53] . In guiding responses towards diverse antigenic stimuli , Blimp1 selectively shifts gene expression profiles in these lineage-committed precursors to directly influence developmental choices and ensure maximally effective protective immunity . It has been known for many years that the Blimp1 consensus motif closely overlaps with the IRF-E sequence recognized by IRF1/IRF2 [20 , 29] . The structure of IRF1 bound to DNA reveals contacts mediated by a conserved cluster of tryptophan repeats [54] . In contrast the first two C2H2 zinc fingers of Blimp1 are required for recognition of its consensus motif [55] . Nonetheless , remarkably these structurally diverse proteins have the ability to interact with the same DNA sequence upstream of important target gene promoters . The present experiments reveal considerable overlap between Blimp1 and Irf1 ChIP-Seq peaks . When further analyzed using Weeder , we identified an IRF-E DNA binding motif GAAAGTGAAA [20] underlying a subset of the 331 overlapping Blimp1/Irf1 ChIP-seq peaks . Of these 7% ( n = 22 ) failed to match the motif , 23% ( n = 75 ) contained a single match , and 70% ( n = 234 ) contained 2 or more IRF-E sequences . For example , the overlapping peak proximal to the Psmb10 promoter contained 8 motif matches . The Tap1 and Psmb8 bi-directional promoter also displays overlapping occupancy [56 , 57] . Closer inspection of the peak profiles reveals two closely adjacent Blimp1 peaks but only one overlaps with Irf1 . The close arrangement of multiple adjacent consensus binding sites potentially allows co-occupancy of Blimp1 and Irf1 at these promoter regions . It is tempting to speculate that simultaneous binding of both repressor ( Blimp1 ) and activator ( Irf1 ) provides a rapid transcriptional switch mechanism . Previous experiments demonstrate that MHC class I surface expression , absent at early embryonic stages of development is interferon-inducible [58 , 59] . Similarly temporal and spatially restricted Irf1 expression is tightly regulated in the early embryo [60] . Irf1 constitutively expressed in the developing thymus plays an essential role in positive and negative selection of the CD8+ T cell repertoire [57 , 61] . MHC class I-restricted CD8+ T lymphocytes are directed against peptides derived from cytosolic proteins degraded by the proteasome that are translocated across the ER membrane by the TAP1/TAP2 transporter , stabilized by interactions with the dedicated class I chaperone Tapasin , and edited by ER aminopeptidases [19] . T cell receptor ( TCR ) repertoire selection is thought to depend on affinity differences as developing T cells interact with self-peptide MHC complexes displayed by thymic antigen-presenting cells in the cortical and medullary compartments [62] . During inflammatory and anti-viral responses the MHC class I peptide loading pathway becomes dramatically up-regulated to activate host defenses [31] . Interestingly both the adult thymus and small intestine constitutively express so-called immunoproteasome subunits [63] . We demonstrate here that numerous IFN-inducible components of the MHC class I antigen presenting pathway are direct Blimp1 targets in the embryonic intestinal epithelium . Irf1 constitutively expressed throughout intestinal development is poised to activate the MHC class I pathway in mature enterocytes coincident with down-regulated Blimp1 expression . In the neonatal intestine however , Blimp1 antagonizes Irf1 target gene expression to prevent premature activation of the MHC class I peptide-loading machinery . Besides its essential role in absorption of nutrients , the intestinal epithelium also function as a protective barrier to prevent infection . The formation of gut-associated lymphoid tissues including the Peyer’s patches is initiated in utero but maturation of the mucosal immune system is incomplete at birth . Neonatal immune tolerance during postnatal colonization by commensal bacteria is essential for the establishment of a well-balanced host-commensal relationship . It is widely accepted that maternal tolerance of the allogenic fetus is largely dependent on the absence of embryonic MHC class I expression [58 , 59] . The present experiments demonstrate that another key feature of mammalian development is competitive binding by Blimp1 and Irf1 at the promoters of IFN-inducible components of the MHC class I machinery in fetal intestine to guarantee immune tolerance during postnatal intestinal maturation . We also identify BLIMP1 expression in human fetal enterocytes , implicating a potentially conserved functional role in guiding neonatal intestinal maturation and metabolic adaption . Although obtaining healthy preterm and neonatal material presents a considerable obstacle , nonetheless our future studies will aim to learn more about how BLIMP1 functions in humans to coordinate neonatal immune tolerance and the developmental switch from immature to mature enterocyte transcriptional programmes . Note added in proof . While this manuscript was under review , another genome wide Blimp-1 ChIP data set was published by Saitou and colleagues ( Kurimoto , K . , Yabuta . Y . , Hayashi , K . Ohta , H . , Kiyonari , H . , Mitani , T , Moritoki , Y . , Kohri , K . , Kimura , H . , Yamamoto , T . , Katou , Y , Shirahige , K . , and M . Saitou ( 2015 ) Quantitative dynamics of chromatin remodelling during germ cell specification from mouse embryonic stem cells . Cell Stem Cell ) [64] . The authors similarly used a Blimp-1-eGFP fusion knock-in strategy , but in contrast to our study , the eGFP cassette was inserted at the N-terminus and the resulting homozygous mice are fertile . They focused exclusively on expression during in vitro PGC specification , so it remains unclear whether germ cell defects in our BEG allele may reflect compromised recruitment of epigenetic partners and/or marginally reduced expression levels . Remarkably despite cell type differences , we see that there is roughly 48% overlap between the Blimp-1 ChIP seq peaks identified here in small intestine and those identified in PGCs by Kurimoto et al . Importantly overlapping targets include Prdm1 , Il33 , Psmb9 , Psmb10 , Psmg4 , Psme2 , Tap1 , Tap2 , Tapbp , and Erap1 .
For the 5’ homology region , a StuI-XbaI fragment was excised from a genomic subclone [6] and introduced into a modified pBluescriptII plasmid ( Stratagene ) . To complete the 5’ homology region PCR ( Platinum Pfx polymerase , Invitrogen ) was performed using the primers Prdm1-KI-F: AGAAACCAGCGCTTCTGTTTTAGTACGCGGAGC and Prdm1-KI-R: GAGAGGCGCGCCGAGAACTAGTCTCTGCCAGTCCTTGAAACTTCACGGAGCC with the bacterial artificial chromosome bMQ-375h16 as template . The PCR product was digested ( XbaI , AscI ) and cloned to introduce SpeI and AscI ( underlined ) cloning sites into Prdm1 exon 3 . The Blimp1-eGFP fusion construct plus SV40 polyadenylation signal [65] excised using XhoI-NotI was subcloned into a pBluscriptII based shuttle vector containing an upstream NheI site and downstream AscI site . This Blimp1-eGFP cDNA expression cassette was then ligated into SpeI and AscI sites in Prdm1 exon 3 . Finally , a fragment containing a Frt flanked Neomycin resistance cassette [66] , the MfeI-SphI 3’ homology region [6] and Hsv-TK cassette were inserted using AscI and PmeI restriction sites . The NotI linearized targeting vector was introduced into CCE embryonic stem ( ES ) cells by electroporation . Southern blot screening was performed as described using the restriction enzyme and probe combinations shown in Fig 1A [27] . Southern probes were amplified by PCR using the following primers: 5'-Forward: GATAGGATCCTTTCCAGCTGTTACTATGTAGG , 5'-Reverse: GATACTCGAGCTTATGCTTCATAGTTAATTTGG , 3'-Forward: GATAGAATTCAATGCCATTTGTCAGGGAGC , 3'-Reverse: GATACTCGAGCTTTTGGCCACAGGACAATG . For excision of the Frt-Neo cassette , correctly targeted ES cell clones were transiently transfected with pCAGGS-FlpO ( generously provided by Bill Skarnes ) and subsequently screened using the 3' probe together with KpnI genomic digest . Genotyping of mice carrying the novel eGFP knock-in ( BEG ) allele was performed with Forward primers BEG Mut ( GTTATTGGCGTGGTAAGTAAGG ) and WT ( AGGCATCCTTACCAAGGAAC ) and Reverse primers BEG Mut ( ATTTATCACTGTGAGCTCTCCAG ) and WT ( GCTGAAGGGAGGAAGAAATG ) . The cycling conditions were 94°C for 20 s , 58°C for 30 s , and 72°C for 45 s for 35 cycles . Conditional mutants selectively lacking Blimp1 function in the developing intestine , generated by crossing Prdm1CA/CA animals to the Villin cre transgenic strain were genotyped as described [16] . The targeted deletion that selectively eliminates exon 1A promoter usage and disrupts Blimp1 dependent plasma cell differentiation as well as the BAC transgenic reporter strain expressing membrane targeted Venus under the control of the Blimp1 regulatory elements have been described [25 , 27] . All animal experiments were performed in accordance with Home Office ( UK ) regulations and were approved by the University of Oxford Local Ethical Committee . Cell cultures to induce plasma cell differentiation and Western blot analysis of LPS-stimulated splenocytes were performed as described previously [65] . To insure quantitative protein recoveries , embryonic and postnatal intestinal tissues were flushed with PBS , snap frozen and immediately dissociated directly in SDS sample buffer . The primary antibodies were: rat monoclonal anti-Blimp1 ( clone 5E7 , SC-130917; Santa Cruz ) , rabbit polyclonal anti-Irf1 rabbit ( M-20 , Santa Cruz , SC-640 ) , rat monoclonal anti-MHC class I ( ER-HR52 , SC-59199; Santa Cruz ) , hamster monoclonal anti-G9a ( clone 14–1 , D141-3; MBL ) , rabbit polyclonal anti-HDAC1 ( AB7028; Abcam ) , rabbit polyclonal anti-HDAC2 ( AB7029; Abcam ) , rabbit monoclonal anti-LSD1 ( clone EPR6825 , AB129195; Abcam ) , rabbit polyclonal anti-CoREST ( 07–455 , Merck Millipore ) and rabbit polyclonal anti-β-tubulin ( SC-9104; Santa Cruz ) . Secondary antibodies were anti-mouse immunoglobulin ( Ig ) –horseradish peroxidase ( HRP ) ( NA931V; GE Healthcare ) , anti-rat Ig—HRP ( NA935V; GE Healthcare ) , anti-rabbit Ig—HRP ( NA934V; GE Healthcare ) , or anti-Armenian hamster Ig—HRP ( SC-2904; Santa Cruz ) . Levels of Blimp1-eGFP fusion protein relative to levels of endogenous Blimp1 protein were quantified on Western blots using a ChemiDoc XRS+ system and Image Lab software ( BIO-RAD ) . Cell staining experiments were performed as described [65] using a FACSCalibur flow cytometer ( BD Biosciences ) , and data were analyzed with FlowJo software ( Tree Star ) . High-resolution images were captured using an ImageStreamX Mk II imaging flow cytometer ( AMNIS ) and analyzed using IDEAS software ( AMNIS ) . Placental and intestinal tissue samples were fixed overnight in 4% paraformaldehyde ( PFA ) in PBS , dehydrated in ethanol , embedded in paraffin and sectioned ( 6 μm ) Immunohistochemistry was performed as described [8] . Primary antibodies were rat monoclonal anti-Blimp1 ( clone 5E7 , SC-130917; Santa Cruz ) , or for human tissue rat monoclonal anti-Blimp1 ( clone 6D3 , 14-5963-82; eBioscience ) , rabbit monoclonal anti-Irf1 ( 8478; Cell signaling ) or rat monoclonal anti-MHC class I ( ER-HR52 , SC-59199; Santa Cruz ) . Visualization of primordial germ cells , by staining for alkaline phosphatase was performed as described [67] . Testes samples were fixed overnight in Bouin’s fixative , dehydrated in ethanol , embedded in paraffin , sectioned ( 6 μm ) and stained with hematoxylin and eosin . Small intestines were dissected , flushed with PBS , cut into small pieces and cross-linked with 1% formaldehyde in PBS for 15 min at 4°C followed by 35 min at 25°C [68] and subsequently processed for ChIP using either 10 μg of mouse anti-GFP IgG2a ( clone 3E6 , A11120; Invitrogen ) , polyclonal rabbit anti-Irf1 antibody ( M-20 , Santa Cruz , SC-640 ) or as a control , normal rabbit IgG ( Santa Cruz , SC-2027 ) as described previously [28] . The DNA samples were multiplexed and sequenced using two lanes on an Illumina HiSeq 2000 sequencer . Duplicate test ( GFP ChIP of Prdm1BEG/BEG , or Irf1 ChIP of WT and Villin-cre conditional Blimp1 mutants , [16] ) samples or individual negative control ( GFP ChIP of wild type or , normal rabbit IgG ChIP ) and input samples were analyzed . Sequence reads were mapped to the mm9 mouse genome release with Stampy using default parameters [69] . Peak calling was performed with MACS1 . 4 . 2 [70 , 71] , using default parameters to call areas of enrichment in ChIP samples over input . Regions of enrichment detected in negative controls samples were removed from subsequent analysis . GFP ChIP peaks called in wild type samples were subtracted from GFP peaks called in BEG/BEG samples . Similarly , normal rabbit IgG ChIP peaks were subtracted from Irf1 peaks . The overlapping peaks in duplicate ChIP samples were then identified and the core region of overlap was further analyzed . The genomic distribution of ChIP-seq peaks compared to gene annotations was determined using CEAS [72] . Genes of Ensembl release 67 with proximal Blimp1 or Irf1 binding were identified using custom Perl scripts . De novo identification of motifs within ChIP-seq peaks was performed using MEME suite tools [73] . Functional annotation of ChIP-seq peaks was performed with GREAT version 2 . 0 . 2 using the basal plus extension rule , annotating genes within 5 kb of transcription start sites initially or within 25 kb when no proximal gene in known to exist [74] . Regions of overlap between the ChIP-seq peaks identified in present study with other published datasets were compared using custom Perl scripts . Peak regions in human datasets were first converted to mm9 using the UCSC liftOver function . The association between Blimp1 binding ( ± 5kb of TSS ) and genes differentially expressed in embryonic Prdm1-/- small intestine [16] ( Gene Expression Omnibus database , www . ncbi . nlm . nih . gov/geo , accession no . GSE29658 ) was calculated by chi-square test . For comparative Irf1 ChIP-seq analysis reads mapping within identified peaks were counted using HTSeq-count [75] . TMM normalization factors and tagwise dispersions were then computed , and differential occupancy between sites in mutant and wild type determined by exact test using edgeR [76 , 77] . False discovery rates were calculated using the Benjamini and Hochberg method . Sites differentially bound with an FDR ≤ 0 . 05 were considered significant . QPCR analysis of triplicate GFP or control IgG ChIP , and input samples of Prdm1BEG/BEG and WT E18 . 5 intestine was performed using QuantiTech SYBR Green master mix ( Q2040143; Qiagen ) on a Rotor-Gene Q ( Qiagen ) . Primers were designed to amplify 100–200 bp regions central to ChIP-seq peak genomic coordinates . Selected genes included c-Myc , Psmb8 , Psmb9 , Tap1 , Tap2 and Tapbp [20 , 24] and several candidates with increased transcript levels in Blimp1 mutants i . e . Prdm1 ( 45 . 3 fold ) , Slc16a5 ( 16 . 1 fold ) , Gsta1 ( 16 . 8 fold ) , 2210407C18Rik ( 15 . 7 fold ) , Trib3 ( 2 . 9 fold ) and Cyp4v3 ( 68 . 9 fold ) [16] . A non-enriched ChIP-seq region in the 3’UTR of the Prdm1 gene was used as a negative control . Primer sequences are shown in S6 Dataset . Fold enrichment of ChIP over input was determined relative to a standard curve generated from log diluted sheared genomic DNA . Total RNA from yolk sac and embryonic small intestinal tissues was isolated using an RNeasy Mini kit ( Qiagen ) , and reverse transcription-PCR ( RT-PCR ) was performed using the OneStep RT-PCR kit ( Qiagen ) as previously described [27] . Primers Ex1AFor and Ex1BFor in combination with Ex3Rev distinguish Prdm1 exon 1A and Prdm1 exon 1B transcripts whereas total Prdm1 transcripts were detected with primers Ex4For and Ex5Rev [27] . The ChIP-seq data have been deposited in NCBI GEO with the accession number GSE66069 .
|
The transcriptional repressor Blimp1/Prdm1 plays a pivotal role in the metabolic switch that occurs in the small intestine during the suckling to weaning transition . Notably , expression profiling of perinatal Blimp1-deficient small intestine revealed premature activation of metabolic genes normally restricted to post-weaning enterocytes . To further elucidate the function of Blimp1 in intestinal development , we engineered a novel Blimp1-eGFP-fusion knock-in mouse strain to perform ChIP-seq analysis . In addition to identifying which metabolic genes are direct Blimp1 targets , ChIP-seq analysis revealed a highly conserved Blimp1/Irf-1 overlapping sites that function to control MHC class I antigen processing during acquisition of neonatal tolerance in the first weeks after birth during early colonization of the intestinal tract by commensal microorganisms . Moreover , immunohistochemical analysis of human fetal intestine suggests that a BLIMP1/IRF-1 axis may also function in human intestinal epithelium development .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Blimp1/Prdm1 Functions in Opposition to Irf1 to Maintain Neonatal Tolerance during Postnatal Intestinal Maturation
|
Myopia is one of the most common ocular disorders worldwide . Pathological myopia , also called high myopia , comprises 1% to 5% of the general population and is one of the leading causes of legal blindness in developed countries . To identify genetic determinants associated with pathological myopia in Japanese , we conducted a genome-wide association study , analyzing 411 , 777 SNPs with 830 cases and 1 , 911 general population controls in a two-stage design ( 297 cases and 934 controls in the first stage and 533 cases and 977 controls in the second stage ) . We selected 22 SNPs that showed P-values smaller than 10−4 in the first stage and tested them for association in the second stage . The meta-analysis combining the first and second stages identified an SNP , rs577948 , at chromosome 11q24 . 1 , which was associated with the disease ( P = 2 . 22×10−7 and OR of 1 . 37 with 95% confidence interval: 1 . 21–1 . 54 ) . Two genes , BLID and LOC399959 , were identified within a 200-kb DNA encompassing rs577948 . RT–PCR analysis demonstrated that both genes were expressed in human retinal tissue . Our results strongly suggest that the region at 11q24 . 1 is a novel susceptibility locus for pathological myopia in Japanese .
Myopia is a refractive error ( http://en . wikipedia . org/wiki/Refractive_error ) of the eye in which parallel rays of light focus in a plane anterior to the retina resulting in blurred vision . Myopia is one of the most common ocular disorders worldwide , and is in much higher prevalence in Asians than in Caucasians . Recent population-based surveys in the elderly reported that the prevalence of myopia was approximately 25% in the Caucasian populations [1] and 40% in the East Asian ( Chinese and Japanese ) populations [2] , [3] . Myopia is divided into two distinct subsets , namely , common and pathological myopia . Pathological myopia , also called high myopia , is distinguished from common myopia , also called low/moderate myopia , by excessive increase in axial length of the eyeball , which is the most important contributor to the myopic refraction [4] , [5] . The axial length of the eyeball in adults is approximately 24 mm , and its elongation by 1 mm without other compensatory changes results in a myopic shift of −2 . 5 to −3 . 0 diopters ( D ) . It has been shown that distribution of the axial lengths of the adult myopic population is bimodal [6] , and the subgroup with elongated axial length in the bimodal distribution corresponds to pathological myopia . This group comprises 1% to 5% of the population [3] , [7] , and is commonly defined by axial length greater than 26 . 0 mm which is equivalent to refractive errors greater than −6 D [8] . The excessive elongation of the eyeball causes mechanical strain with subsequent degenerative changes of the retina , choroid , and sclera . The degenerative changes at the posterior pole of the eye such as chorioretinal atrophy or posterior staphyloma are clinically important and unique to pathological myopia [9] . These unique degenerative changes at the posterior pole result in uncorrectable visual impairment due to decreased central vision and make pathological myopia one of the leading causes of legal blindness in developed countries [10]–[13] . It has been reported that not only environmental factors , such as near work and higher education , but also genetic factors contribute to the development of myopia , in particular , of pathological myopia [14] . Previous twin studies reported that the estimated heritability of refractive error and axial length is up to 0 . 90 [15] , [16] , although that might be overestimated due to common environmental effects [17] . Multiple family-based whole genome linkage analyses of myopia reported at least 16 susceptible chromosomal loci ( MYP1–16 in OMIM database; 10 loci for pathological myopia [18]–[27] and 6 for common myopia [28]–[30] ) . Among them , at least 8 chromosomal loci , such as 12q21–23 ( MYP3 ) , 22q12 ( MYP6 ) and 2q37 . 1 ( MYP12 ) were successfully validated by at least two independent studies [31] , [32] . However , no genes responsible for the disease have been identified . The genome-wide association ( GWA ) study using single nucleotide polymorphisms ( SNPs ) as markers is an alternative approach to identify genetic risk factors of common diseases . This approach has been successfully applied to identify genetic risk factors for multigenetic diseases including ophthalmic diseases such as age-related macular degeneration [33] , [34] and exfoliation syndrome [35] . To identify the genetic risk factors of pathological myopia , we conducted a two-stage GWA-based case/control association analysis using 411 , 777 markers with 830 Japanese patients and 1 , 911 Japanese controls ( 297 cases and 934 controls in the first stage , and 533 cases and 977 controls in the second stage ) .
A total of 839 pathological myopic patients with axial length greater than 26 . 0 mm in both eyes were enrolled in the current study . In order to maximize the detection power , patients with axial length greater than 28 . 0 mm in both eyes were enrolled in the first stage of genome scan . No other clinical features were accounted for the assignment of patients to either stage . 824 out of 839 patients ( 98 . 2% ) had degenerative changes specific to pathological myopia . Other features of cases and controls who passed quality control procedures of genotyping results ( see Materials and Methods ) were summarized in Table 1 . For the first stage , we scanned the genome of 302 cases using the Illumina HumanHap550 BeadChip , which launches 561 , 466 relatively frequent SNPs ( minor allele frequency>0 . 05 ) distributed across the human genome at an average interval of 6 . 5 kilobases ( kb ) . Five cases and 149 , 689 SNPs were excluded due to quality control criteria ( see details in Materials and Methods ) and genotyping results of 411 , 777 SNPs in autosomes for 297 cases were used for the statistical analysis . They were compared with 934 controls from the JSNP database [36] for association with phenotype using χ2 test for trend . Genomic Control ( GC ) method [37] revealed only a slight inflation of the test statistics ( GC parameter λ = 1 . 068 ) . We identified 29 SNPs in 22 chromosomal regions with P-value adjusted by GC being smaller than 10−4 ( Figure 1 and Table S1 ) . Among them , seven SNPs at chromosome 8p12 were in strong linkage disequilibrium ( LD ) and likewise two SNPs at chromosome 10q22 . 2 ( pair-wise D′>0 . 95 and r2>0 . 9 ) . Thus , we selected one representing SNP from each region and tested 22 SNPs in the second stage . For the second stage analysis , 537 cases and 980 population controls were genotyped by Taqman method . Among them , four cases and three controls were excluded due to low call rates ( <90% ) . Genotyping success rates of the 22 SNP markers in the remaining 1 , 510 samples were greater than 96 . 8% . The genotype counts of the first and second stages were combined for meta-analysis . One SNP , rs577948 , showed a strongly suggestive association ( P = 2 . 22×10−7 ) ( Table 2 ) in the meta-analysis whereas the remaining 21 SNPs were not significant ( P>10−5 ) ( Table S1 ) . The SNP rs577948 which showed P = 2 . 22×10−7 by meta-analysis with OR of 1 . 37 ( 95% confidence interval ( CI ) : 1 . 21–1 . 54 ) for the risk allele ( nominal P = 2 . 80×10−5 and P = 1 . 42×10−3 in the first and second stages , respectively ) ( Table 2 ) was located at chromosome 11q24 . 1 ( Figure 2A ) . Using the results of the first stage , an LD block which extended a 55-kb region containing rs577948 was generated . Six additional SNP markers within the block were included in the genome scan chip ( Figure 2B ) . Among them , we selected three markers with adjusted P-value smaller than 0 . 01 in the first stage for further genotyping by Taqman method with DNAs used for the second stage . Weaker associations than that of rs577948 were obtained for these three markers by meta-analysis ( Table 2 ) . As shown in Figure 2B , two genes were located in a 200-kb region containing rs577948 . BLID is a cell death inducer containing BH3-like motif [38] , which is located approximately 44-kb upstream of rs577948 . The other gene , LOC399959 , is a hypothetical non-coding RNA [39] which encompassed 114-kb DNA in the region , and rs577948 is located in its second intron . BLID is known as a cell-death inducer expressed in cytoplasm , in mitochondria at lower abundance , and in various human cancer cells from different tissues [38] . LOC399959 was reported as a hypothetical non-coding RNA with a relatively ubiquitous expression pattern . We assessed the expression of the genes by RT-PCR using cDNAs of human retina and brain and those of HeLa cells as positive control . Expressions of both genes were detected in human retinal tissue as well as in human brain and HeLa cells ( Figure 3 ) .
Myopic refraction and axial length are reported to be a complex trait under polygenic control in which contribution of each gene is relatively small [40] . In the current study , two-stage GWA analysis identified a region at chromosome 11q24 . 1 , in which rs577948 showed strongly suggestive P = 2 . 22×10−7 with OR of 1 . 37 ( 95% CI: 1 . 21–1 . 54 ) for the allele G . Our GWA study identified only one strongly-suggestive locus . This may principally be due to the sample size of our study not being adequate . Recent genetic studies of complex traits with higher prevalence enroll much larger number of samples . In contrast , recruitment of patients with pathological myopia is difficult due to its lower prevalence , particularly those with degenerative changes ( namely degenerative myopia ) . In order to improve insufficient detection power , we assigned pathological myopia patients with longer axis ( greater than 28 . 0 mm ) to the first stage . This strategy might be the reason we were successful in identifying the candidate region with relatively small number of cases . Insufficiency of detection power due to a limitation in sample number may be a reason for difference between the findings of preceding linkage studies and ours . OMIM database lists 10 MYP regions ( MYP1–5 , 11–13 , 15 and 16 ) for pathological myopia [18]–[27] and 6 MYP regions ( MYP6–10 and 14 ) for common myopia [28]–[30] . None of these 16 MYPs are on chromosome 11q . Stambolian and colleagues reported heterogeneity LOD score of 1 . 24 at 11q23 in their linkage study for common myopia in Ashkenazi Jewish descent , which is the closest locus to our region reported to date [29] . Because the linkage signal was not strong and the band 11q23 ( chr11 , position 110 , 000 kb to 120 , 700 kb in the NCBI database ) is more than 800 kilobases apart from our LD block in 11q24 . 1 ( chr11 , position 121 , 535 kb to 121 , 590 kb ) , whether or not they overlap each other is inconclusive . On the other hand , our study did not identify the associated SNPs in any of MYPs . Although the insufficiency of detection power may be a reason for difference between our study and the linkage studies , there are other possible reasons . In general , any difference in the study designs could cause heterogeneous results . Firstly , there are two definitions of pathological myopia based on two distinct criteria , namely , the axial length and refractive error . In the current study , we enrolled pathological myopic patients based on the axial length ( greater than 26 . 0 mm in both eyes ) , and not on the refractive error commonly used in the previous studies ( refractive errors greater than −6 D ) . We focused on patients with vision-threatening degenerative changes [9] and the axial length fits better than refractive error for our purpose . The mean refraction in our myopic patients was −13 . 14±4 . 57 D ( eyes that had undergone cataract surgery or corneal refractive surgery were excluded from this calculation ) which indeed correspond to pathological myopic group in the previous linkage studies . On the other hand , it is not clear whether the patients enrolled in the linkage studies fulfill our criteria because the distribution of axial length and degenerative phenotypes in the cases are unknown . The difference in definition of pathological myopia may result in different susceptibility loci between studies . Secondly , the methodology used is different between studies , namely , linkage analysis and association analysis using linkage disequilibrium mapping . The results of linkage and association studies of complex genetic traits are often different . Family-based linkage analysis is much more suitable for identifying rare genetic variants with large effects whereas SNP-based GWA analysis is more powerful in detection of relatively common variants with smaller effects in complex diseases [41] . Finally , the difference can also be due to the ethnicities of the samples enrolled . In the current study , all cases and controls were Japanese . Only one genome-wide linkage study has previously been published for pathological myopia in Japanese [42] and the others were for non-Japanese populations . It would be interesting and important to examine the association of our locus in other ethnicities . Ethnic variations in disease susceptibility genes have been reported in various genetic traits including ophthalmological disorders . One such example is an SNP in the complement factor H gene ( rs1061170 ) which has a large effect size with age-related macular degeneration in Caucasians [33] , [43] , [44] but much smaller in East Asian populations due to a remarkably lower risk allele frequency ( ∼35% in Caucasians and ∼5% in East Asians ) [45] . Another example is exfoliation syndrome and LOXL1 where the risk allele of rs1048661 is inverted between Icelandic ( allele G ) and Japanese ( allele T ) populations [35] , [46] . Because of a large variation in prevalence of myopia among ethnic groups , a future trans-ethnic investigation of myopia risk genes will be important to dissect genetic backgrounds underlying the etiology of myopia . Although the susceptibility locus contains BLID and LOC399959 , it seems premature to discuss the involvement of LOC399959 in myopia since it is a hypothetical non-coding gene . BLID plays a proapoptotic role involving the BH3-like domain by inducing a caspase-dependent mitochondrial cell death pathway [38] . Indeed , several animal and pathological studies suggested the functional role of apoptosis in pathological myopia [47] , [48] . Moreover , a recent genome-wide linkage study followed by a fine-scale association mapping identified a myopia susceptibility gene locus containing the PARL gene which inhibits the mitochondrial pathway of apoptosis by interaction with OPA1 [49] . In this context , BLID seems functionally relevant with the pathogenesis of pathological myopia . However , the true functional origin of association in this region has yet to be determined by further detailed investigation along with replication studies to validate our findings .
All procedures used in this study conformed to the tenets of the Declaration of Helsinki . The Institutional Review Board and the Ethics Committee of each institution approved the protocols used . All the participants were fully informed of the purpose and procedures , and a written consent was obtained from each . Japanese pathological myopic cases were recruited at the Center for Macular Diseases of Kyoto University Hospital , the High Myopia Clinic of Tokyo Medical and Dental University , and Fukushima Medical University Hospital . All subjects underwent comprehensive ophthalmologic examinations , including dilated indirect and contact lens slit-lamp biomicroscopy , automatic objective refraction evaluation , and measurement of the axial length by applanation A-scan ultrasonography ( UD-6000 , Tomey , Nagoya , Japan ) or partial coherence interferometry ( IOLMaster , Carl Zeiss Meditec , Dublin , CA ) . As a general population control of the first stage , genotype count data of 934 healthy Japanese subjects were obtained from the JSNP database [36] . For the second stage , 980 healthy Japanese individuals were recruited at Aichi Cancer Center Research Institute . Genomic DNAs were extracted from peripheral blood leukocytes with QuickGene-610L DNA extraction kit ( FUJIFILM Co . , Tokyo , Japan ) . We designed to scan the genome in two stages . A total of 839 patients and 1 , 914 controls were separated into two groups; 302 cases and 934 controls for the first stage , and 537 cases and 980 controls for the second stage . In order to increase the detection power , patients with longer axis of the eyeball ( greater than 28 . 0 mm ) were principally assigned to the first stage . For the first stage analysis , 561 , 466 SNPs were genotyped in 302 patients of pathological myopia using Illumina HumanHap550 chips ( Illumina Inc . , San Diego , CA ) . This chip covers approximately 87% of the common genetic variations in the Asian population [50] . Cluster definition for each SNP was performed using Illumina BeadStudio Genotyping Module . A systematic quality control procedure of the genome scan results was applied as follows . Samples were evaluated for data quality first and markers were subsequently excluded . Genetic proximity of sample pairs was evaluated with pi-hat in PLINK [51] and four samples with indication of kinship or sample duplication were excluded . Genotypes in X chromosome were used for checking the precision of the phenotype record , and only one sample was removed due to mismatch in gender . The final sample size of pathological myopia was 297 . As a population-based control , genotype count data by the genome scanning of 934 healthy Japanese subjects using the same chip were obtained from the JSNP database [36] . The chip contained 515 , 154 markers in autosomes that are common in the cases and controls . We excluded 78 SNPs due to low successful call rate ( <95% ) in the cases , 1 , 760 SNPs due to the distortion of Hardy-Weinberg Equilibrium ( HWE ) in the controls ( P<10−3 by HWE exact test ) and 46 , 722 monomorphic SNPs . 54 , 817 SNPs with minor allele frequency less than 0 . 05 in both cases and controls were also excluded . After these quality control procedures , a total of 411 , 777 SNPs were used for the statistical analysis . The genotyping call rate was greater than 97 . 43% ( median call rate 99 . 99% ) for DNA sample and 98 . 21% ( median call rate 100% ) for SNP marker . Association between genotypic distribution of each SNP and the disease was examined using a χ2 test for trend . The OR and the 95% CI were estimated using Woolf's method [52] . Inflation in the test statistics was assessed using the genomic-control method [37] . Haploview [53] software was used to infer the LD in the targeted regions . SNPs with P-value adjusted by genomic control being smaller than 10−4 were selected as candidates for second stage . Among the candidate SNPs , LD indices ( D′ and r2 ) were calculated with Haploview and when multiple SNPs were in strong LD ( D′>0 . 95 and r2>0 . 9 ) , one representative SNP was chosen to be genotyped in the second stage . In the second stage , 537 cases and 980 controls were genotyped with the Taqman SNP assay using the ABI PRISM 7700 system ( Applied Biosystems , Foster City , CA ) . The 302 pathological myopic cases in the first stage were also genotyped to validate the concordance between Illumina Infinium assay and Taqman assay . Samples with low successful call rate ( <90% ) were excluded from the study . Subsequently four cases and three controls were excluded and data of 533 cases and 977 controls were used for the analysis . The concordance rate ranged between 98 . 68% and 100% for the 22 SNPs . The genotype counts of the first and second stages were combined for meta-analysis using the Mantel-Haenzel method [54] as a fixed-effect model . The OR heterogeneity between the first stage and the second stage was evaluated using Cochran's Q-statistic P-value . The data from the second stage were also evaluated for association independently from the first stage . Human retina cDNAs were obtained from Takara Bio Inc . ( Kyoto , Japan ) . Total RNA of HeLa cells and human whole brain were also obtained from the same manufacturer and cDNAs were synthesized using the First-Strand cDNA Synthesis Kit ( GE Healthcare Life Sciences , Piscataway , NJ ) . Two pairs of oligonucleotides were synthesized for RT-PCR; 5′-TTGGGTTCCAACAAAGAACC-3′ and 5′-CTTTTACAGGGCCTCAGCAG-3′ for BLID , and 5′-GGCGACATCAGACAGACAGA-3′ and 5′-AGGACCAGCTGAAAGGAACA-3′ for LOC399959 . Expression of glyceraldehyde-3-phosphate dehydrogenase ( GAPDH ) was tested for cDNA quantification using 5′-GACAACAGCCTCAAGATCATCA-3′ and 5′-GGTCCACCACTGACACGTTG-3′ . PCR reactions were performed under the following condition: initial denaturation at 96°C or 2 minutes , followed by 35 cycles ( for BLID and LOC399959 ) or 18 cycles ( for GAPDH ) at 96°C for 20 seconds , 60°C for 40 seconds , and polymerization at 72°C for 40 seconds .
|
Myopia is one of the most common ocular disorders with elongation of axis of the eyeball . Pathological myopia or high myopia , a subset of myopia which is characterized with excessive axial elongation and degenerative changes of the eye , is a leading cause of visual impairment . Since genetic factors play significant roles in its development , identification of genetic determinants is an urgent and important issue . Although family-based linkage analyses have isolated at least 16 susceptible chromosomal loci for pathological or common myopia , no gene responsible for the disease has been identified . We conducted the first genome-wide case/control association study of pathological myopia in a two-stage design using 411 , 777 markers with 830 Japanese patients and 1 , 911 Japanese controls . We identified a region strongly suggestive for the disease susceptibility at chromosome 11q24 . 1 containing BLID and LOC399959 . Their expression was confirmed in human retina with RT–PCR . BLID encodes an inducer of apoptotic cell death , and apoptosis is known to play an important functional role in pathological myopia . We believe that our study contributes to further dissect the molecular events underlying the development and progression of pathological myopia .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"ophthalmology",
"genetics",
"and",
"genomics/complex",
"traits",
"ophthalmology/retinal",
"disorders"
] |
2009
|
A Genome-Wide Association Analysis Identified a Novel Susceptible Locus for Pathological Myopia at 11q24.1
|
Innate immune restriction factors represent important specialized barriers to zoonotic transmission of viruses . Significant consideration has been given to their possible use for therapeutic benefit . The apolipoprotein B mRNA editing enzyme catalytic polypeptide 3 ( APOBEC3 ) family of cytidine deaminases are potent immune defense molecules capable of efficiently restricting endogenous retroelements as well as a broad range of viruses including Human Immunodeficiency virus ( HIV ) , Hepatitis B virus ( HBV ) , Human Papilloma virus ( HPV ) , and Human T Cell Leukemia virus ( HTLV ) . The best characterized members of this family are APOBEC3G ( A3G ) and APOBEC3F ( A3F ) and their restriction of HIV . HIV has evolved to counteract these powerful restriction factors by encoding an accessory gene designated viral infectivity factor ( vif ) . Here we demonstrate that APOBEC3 efficiently restricts CCR5-tropic HIV in the absence of Vif . However , our results also show that CXCR4-tropic HIV can escape from APOBEC3 restriction and replicate in vivo independent of Vif . Molecular analysis identified thymocytes as cells with reduced A3G and A3F expression . Direct injection of vif-defective HIV into the thymus resulted in viral replication and dissemination detected by plasma viral load analysis; however , vif-defective viruses remained sensitive to APOBEC3 restriction as extensive G to A mutation was observed in proviral DNA recovered from other organs . Remarkably , HIV replication persisted despite the inability of HIV to develop resistance to APOBEC3 in the absence of Vif . Our results provide novel insight into a highly specific subset of cells that potentially circumvent the action of APOBEC3; however our results also demonstrate the massive inactivation of CCR5-tropic HIV in the absence of Vif .
Innate immune restriction factors embody specialized barriers to zoonotic transmission of viruses . Substantial consideration has been given to their potential use for therapeutic benefit [1] , [2] . The apolipoprotein B mRNA editing enzyme catalytic polypeptide 3 ( APOBEC3 ) family of cytidine deaminases are potent innate immune defense factors capable of efficiently restricting endogenous retroelements as well as a diverse range of viruses including Hepatitis B virus , Human Immunodeficiency virus , Human T Cell Leukemia virus , TT virus , and Human Papilloma virus [3]–[8] . The best-characterized APOBEC3 family members are the immune defense molecules APOBEC3G ( A3G ) and APOBEC3F ( A3F ) and their lethal restriction of HIV [5] , [9] . HIV has evolved to counteract these powerful restriction factors by encoding an accessory gene designated viral infectivity factor ( vif ) . In vitro studies have elegantly shown that in the absence of Vif , A3G and A3F are encapsidated into nascent virions and deaminate cytosines in the minus strand of HIV DNA during reverse transcription [10]–[12] . APOBEC3 deamination of cytosines in the minus strand of the viral genome occurs at both CC and TC dinucleotide sites , resulting in GG to AG as well as GA to AA mutations in the coding strand of the viral genome [10] , [11] , [13] , [14] . APOBEC3 induced G to A mutations at GG dinucleotide sites are exclusively the result of A3G deamination , while mutations occurring at GA sites can be caused by multiple APOBEC3 proteins including both A3F and A3G [10] , [15] . While studies have demonstrated the deleterious effects of G to A hypermutation of the HIV genome [10] , [16]–[18] , a recent in vitro study showed variable levels of A3G induced G to A mutations suggesting that A3G may contribute to viral diversity [19] . In this study , we use humanized mice for the in vivo study of HIV in the context of a human immune system . Both NSG-hu and NSG BLT mice are systemically reconstituted with multiple lineages of hematopoietic cells including T cells , B cells , and myeloid cells following transplantation with CD34+ hematopoietic stem cells [20] , [21] . Additionally , BLT humanized mice are implanted with human liver and thymic tissue under the kidney capsule prior to the transplant of autologous CD34+ cells which results in the development of a bona fide human thymus for T cell development [21] . Like any other model for HIV/AIDS research humanized mice have several strengths and limitations that have to be taken into consideration in the development of experimental plan . Two recent review articles cover this area in significant detail [22] , [23] . Despite their limitations humanized mouse models have previously been used for the study of HIV transmission , pathogenesis , prevention , therapy and latency/eradication [20] , [24]–[28] . Here we first demonstrate the highly effective inactivation of CCR5-tropic HIV-1 by APOBEC3 when unobstructed by a functioning vif in vivo after intravenous infection . Secondly , we demonstrate that if injected directly into the thymus , vif-defective viruses can replicate escaping absolute APOBEC3 restriction .
To confirm that the mutations disrupting vif do not have a detrimental effect on the replicative capacity of HIV-1JR-CSF , we generated a CCR5 expressing permissive cell line ( CEM-SS CCR5 ) and infected them with wild-type HIV-1JR-CSF or isogenic viruses containing either an irreparable deletion in vif ( HIVJR-CSFΔvif ) or a one base insertion in vif ( HIVJR-CSFvifFS ) . Replication of both vif-defective viruses was equal to that of wild-type in permissive cells , confirming that the disruption of vif did not have a deleterious effect on HIV-1 in the absence of APOBEC3 ( Figure 1A ) . To assess the in vivo effectiveness of APOBEC3 restriction of HIV , we intravenously infected NSG-hu mice with wild-type HIV-1JR-CSF ( a T-cell CCR5-tropic primary isolate ) or HIVJR-CSFΔvif . As early as one week after intravenous infection , widespread replication of wild-type virus was detected as HIV DNA was amplified from every tissue analyzed ( Figure 1B ) . In contrast the vif-defective virus did not sustain replication in humanized mice; as viral DNA was sparsely present ( Figure 1B ) . Notably , HIVJR-CSFΔvif DNA could only be amplified from one organ from each infected mouse suggesting that an extremely low number of infected cells were present . Analysis of the viral DNA sequence from the animals revealed that HIV DNA from mice infected with wild-type virus had no mutations , whereas viral DNA from the HIVJR-CSFΔvif infected mice had numerous G to A mutations consistent with APOBEC3 induced restriction ( Figure 1C ) . The limited number of tissues with cells harboring G to A mutated HIVJR-CSFΔvif DNA one week after exposure suggests that APOBEC3 restriction of vif-defective HIV occurs rapidly in vivo . While evidence of APOBEC3 restriction of vif-deficient HIV is observed early after infection , we next determined whether HIVJR-CSFΔvif could develop resistance to APOBEC3 and replicate systemically . To address this , we infected humanized mice ( n = 8 ) intravenously with HIVJR-CSFΔvif and monitored them for plasma viral load . Longitudinal analysis demonstrated that HIVJR-CSFΔvif restriction by APOBEC3 is absolute , as no viral RNA was present in the plasma of the mice at any time point in contrast to infection with wild-type HIVJR-CSF ( Figure 2A ) . No revertants or complementary changes arose that restored the ability of HIVJR-CSFΔvif to replicate . In contrast to the widespread presence of viral DNA in all tissues analyzed after wild-type HIV infection , the extremely limited replication of HIVJR-CSFΔvif was confirmed by the absence of HIV DNA in tissues obtained from 4/8 infected mice , and the presence of lethally mutated viral DNA in only a few tissues of the other four animals ( Figure 2B ) . To determine the extent of APOBEC3 hypermutation in the HIVJR-CSFΔvif provirus , we analyzed the sequences for G to A mutations at GG , GA and GY dinucleotide sites which are the targets of the APOBEC3 proteins . We found that 25–65% of the GG sites had been mutated with additional ( albeit fewer ) mutations present at GA sites , demonstrating extensive APOBEC3 hypermutation to lethally restrict HIVJR-CSFΔvif ( Figure 2C ) . Analysis of the mutational profile in the HIVJR-CSFΔvif DNA from the mice showed that 84% of all the G to A mutations occurred at GG dinucleotide sites whereas only 15% of mutations were present at GA sites and only 1% occurred at GY sites ( Figure 2D ) . Taken together , these results demonstrate that HIVJR-CSFΔvif is unable to overcome the loss of Vif and is lethally restricted by APOBEC3 in vivo . To evaluate the selective pressure exerted by APOBEC3 on HIV in vivo , we used a mutant isogenic virus containing a one base insertion in vif ( HIVJR-CSFvifFS ) to intravenously infect 16 humanized mice representing 7 different human donors . Consistent with our previous results , in 10/16 mice intravenously infected with HIVJR-CSFvifFS there was no evidence of virus replication as determined by the absence of viral RNA in the plasma ( Figure 3A ) . However , in the remaining six mice the virus was able to replicate to levels similar to those observed with the wild-type virus ( Figure 3A ) . One salient aspect noted was the almost complete absence of APOBEC3 mutations in viral RNA samples obtained from the plasma from these six mice . Molecular analysis of viral sequences from the peripheral blood from these mice demonstrated that in all six cases a one-nucleotide deletion had occurred that fully restored the vif open reading frame ( ORF ) highlighting the extreme selective pressure APOBEC3 exerts on HIV in vivo to restore Vif activity or be lethally mutated . Two important issues that should be noted are 1 ) the virus used for these experiments was generated via transient transfection of 293T cells creating a uniform inoculum and 2 ) that this repair mutation occurring in vivo is not at a putative APOBEC3 site and therefore it is most likely is a result of a mutation occurring during reverse transcription . We then analyzed viral DNA from the tissues of all 16 mice exposed to HIVJR-CSFvifFS . In multiple tissue samples from four of the aviremic mice we found no evidence of HIV DNA . In similar samples from six other aviremic mice only low levels of heavily mutated HIV DNA was present in a few tissues ( 30–60% of the GG dinucleotide sites mutated ) ( Figure 3B and 3C ) . The mutational profile of the viral DNA from these animals again showed a preference for GG sites accounting for 87% of all mutations ( Figure 3D ) . In sharp contrast , in mice where the vif ORF was restored virtually intact viral DNA was present in every tissue analyzed , highlighting the strong selective pressure exerted by APOBEC3 on HIV ( Figure 3B and 3C ) . The lack of G to A mutations in the vif-restored viral genome suggested that the virus had evaded APOBEC3 restriction until restoring vif . The stochastic nature of vif ORF restoration may reflect its occurrence in a specific anatomical location ( s ) . Therefore we directly injected 9×104 TCIU of HIVJR-CSFvifFS into the spleen , liver , lung or human thymic implant of separate humanized mice . Evidence of viral replication in peripheral blood was exclusively found when HIVJR-CSFvifFS was injected directly into the thymus ( Figure 4A ) . In this case , sequence analysis of vif showed a one base deletion restoring Vif expression . Injection of the virus into the spleen , liver , or lung resulted in absolute restriction with no viremia and no residual viral DNA present in any tissue ( Figure 4A and 4B ) To determine if restoration of the vif ORF following thymic injection was non-random , we increased the virus inoculum four-fold and repeated the infection , using 3 . 6×105 TCIU of HIVJR-CSFvifFS injected into the same set of tissues . Again evidence of HIV replication was only observed after intrathymic exposure with 3/3 mice that received the virus directly into the thymus becoming viremic ( Figure 4A ) . Strikingly , despite the high virus inoculum injected directly into the spleen , liver , or lung no evidence of virus replication was observed ( Figure 4A ) . When tissues from these animals were analyzed , we found that only one mouse , FS24 , had viral DNA present ( Figure 4B ) and that it had all been lethally mutated by APOBEC3 ( Figure 4C and 4D ) . These results demonstrate that transient HIVJR-CSFvifFS replication and subsequent vif restoration specifically occurs following a direct thymic exposure . Furthermore , the potent antiretroviral activity of APOBEC3 is highlighted by the absolute restriction of HIVJR-CSFvifFS when the virus is injected into other tissues . Since the reversion of the vif ORF specifically occurred following injection of the virus into the thymus , we next determined whether A3G and A3F expression was lower in the thymus compared to other tissues . We tested this by comparing A3G and A3F mRNA levels in purified thymocytes ( of which >90% are CD4+ ) with those in CD4+ cells isolated from other tissues in humanized mice . Our results show that thymocytes express 4–8 fold less A3G mRNA and 2 . 5–3 . 5 fold less A3F mRNA than human CD4+ cells isolated from the spleen , liver or lung ( Figure 5A and Figure S1 ) . Furthermore , no difference in A3G or A3F mRNA expression was found in thymocytes from humanized mice or human thymus . Additionally , A3G in thymocytes was found to be 3–4 fold lower compared to human peripheral blood mononuclear cells ( PBMC ) by both mRNA and protein expression ( Figure 5A and 5B ) . These results are consistent with the observation that vif reversion specifically occurs following thymic injection of HIVJR-CSFvifFS and suggests that the thymus may support Vif-independent HIV replication . Based on our observations of reduced A3G and A3F expression in thymocytes and reversion of the vif ORF exclusively occurring with thymic exposure , we considered the possibility that a direct injection of HIVJR-CSFΔvif ( the virus containing a non-revertible deletion in vif ) into the thymus would result in Vif-independent replication . HIV RNA was transiently observed in the plasma of 2/6 mice following thymic infection with this virus ( Figure S2 ) . This low level of replication in some mice is consistent with the results presented above with frame shift containing HIVJR-CSFvifFS ( Figure 4A ) , in which the virus had restored the vif ORF and was able to replicate unimpeded by APOBEC3 after reversion . These results show that recovery of Vif activity is necessary for ongoing replication and viral dissemination by vif-defective HIVJR-CSF . We hypothesized that the lack of robust and sustained replication of HIVJR-CSFΔvif following direct thymic infection could be due to limited CCR5 expression in the thymus , as <5% of thymocytes express CCR5 whereas 30–40% of thymocytes express CXCR4 [29]–[31] . We therefore introduced the deletion described above into the vif ORF of HIV-1LAI , a CXCR4-tropic virus ( HIVLAIΔvif ) and confirmed that the disruption of vif did not affect the ability of the virus to replicate in the absence of APOBEC3 ( Figure S3 ) . We tested our hypothesis by directly injecting HIVLAIΔvif into the thymus of four humanized mice . Viremia was present in 4/4 animals inoculated in this manner ( Figure 6A ) . In contrast , when HIVLAIΔvif was directly injected into the spleen , liver , or lung of an additional 3 animals viral replication was absolutely restricted ( Figure 6A ) . These results further demonstrate that Vif-independent HIV replication can be sustained following exposure into the thymus but is vigorously restricted in other tissues . Additionally , when HIVLAIΔvif was injected intravenously , sustained levels of viral replication were observed in the plasma of humanized mice; however this replication was lower relative to the parental virus ( Figure 6B ) . Remarkably , unlike wild-type HIVLAI which rapidly depletes peripheral blood CD4+ T cells , infection with HIVLAIΔvif did not deplete CD4+ T cells in the periphery despite sustained viral replication ( Figure 6C ) . Sequencing of viral RNA obtained from the plasma of HIVLAIΔvif infected mice showed significantly fewer G to A mutations compared to the same region of viral DNA isolated from PBMC , suggesting that the infection was likely being sustained in cells with lower APOBEC3 expression ( Figure 6D ) . Consistent with these results , HIVLAIΔvif DNA was abundant in the tissues of intrathymically or intravenously exposed mice while direct exposure into the spleen , liver , or lung resulted in viral DNA sparsely present in the organs ( Figure 6E ) . Since this virus could not restore Vif expression , its viral DNA had G to A mutations; however , consistent with the low expression of A3G and A3F in the thymus ( Figure 5A , 5B and S1 ) , significantly fewer G to A mutations were present in viral DNA amplified from the thymus when compared to the same region of the viral DNA amplified from other organs ( Figure 6F ) . Analysis of the mutational profile in the HIVLAIΔvif DNA from the mice showed that 86% of all the G to A mutations occurred at GG dinucleotide sites ( Figure 6G ) . The presence of hypermutated provirus in several tissues suggests that the HIV was not able to develop resistance to APOBEC3 by second site mutations in the absence of vif , but was instead persisting in a pool of cells that permitted replication ( Figure 7 ) . Taken together , these results demonstrate that HIV can sustain replication independent of vif escaping APOBEC3 restriction in vivo .
Our results demonstrate evidence of vif-independent replication of HIV in an in vivo setting . The observation that infection with CCR5 tropic HIV is rapidly extinguished in the absence of vif while CXCR4 tropic HIV lacking vif can sustain replication suggests that vif-independent HIV replication is occurring in a location with a paucity of cells expressing CCR5 , such as the thymus where <5% of cells express CCR5 while a far greater number of cells ( 30–40% ) express CXCR4 [29]–[31] . Consistent with this observation , direct injection of vif-deficient HIV into organs resulted in detectable viral replication only following thymic infection . Remarkably , reversion of the vif ORF with HIVJR-CSFvifFS occurred in 100% of thymic exposures while being absolutely restricted when injected into all other organs , highlighting the potent antiretroviral activity of APOBEC3 . Interestingly , despite the strong selective pressure applied by APOBEC3 , we did not observe any evidence of vif-defective HIV-1JR-CSF altering its coreceptor usage to take advantage of the lower A3G and A3F expression in the thymus . Coreceptor switching is a complicated process involving multiple mutations in envelope that occurs over a period of years in patients [32] . During their short lifespan , coreceptor switching is not common in humanized mice and has only been reported in a single mouse [33] . Our results demonstrate massive inactivation of CCR5-tropic HIV-1 when the protective effects of Vif are absent . These results are consistent with previously published work by Sato et . al . [34] . Under their experimental conditions , these investigators found that vif-defective CCR5-tropic HIV did not replicate at all in humanized mice . Replication of HIV-1 with a functional vif gave a different result . In this case , there was a low level of G to A mutation in both A3G and A3F contexts in viral DNA sequences [34] . Thus , these authors confirmed in the humanized mouse model the early observations of the occasional occurrence of hypermutation of HIV-1 isolated from patients [35]–[38] . Analysis of HIV DNA in aborted infections for G to A hypermutation , the hallmark of APOBEC3 restriction , demonstrated that when HIV DNA was present , there was an overwhelming prevalence of mutations at GG dinucleotide sites indicating that in the absence of vif A3G is the dominant HIV restricting factor in vivo [10]–[12] . This conclusion is further supported by two recent papers using stably expressed A3F or gene targeting to create null mutants to systematically disrupt the individual APOBEC3 proteins that have elegantly demonstrated that A3G is the APOBEC3 family member that induces the preponderance of GG to AG mutations in vif-deficient HIV DNA [15] , [39] . One substantial benefit of A3G restriction is that the mutation of GG to AG can be highly effective in inactivating viral genes because of the conversion of tryptophan codons ( TGG ) to stop codons ( TAG , TGA , or TAA ) . The lower level of GA to AA mutations that we observed suggests a contributory role for A3F in the overall level of G to A mutations we observed . The impact of A3F remains unclear however since the specificity of A3G for the GG context is not absolute and some of the GA to AA mutations we observed may have been created by A3G . A role for A3F in HIV restriction has been questioned recently but this issue remains unresolved in vivo [39] , [40] . Future experiments with humanized mice will address this question . The results presented here demonstrate that in vivo HIV fails to develop second-site mutations to compensate for the absolute loss of vif to overcome A3G induced mutation , which is in contrast to observations made with in vitro systems with ectopically expressed A3G [41] , [42] . This potent restriction of HIV in vivo is not observed by inactivation of other HIV-1 accessory genes [43] , [44] . To survive in vivo in the absence of vif , HIV relies on target cells with reduced A3G expression in which it can replicate as shown by the lack of G to A hypermutation in the cell free virus despite the abundance of G to A mutations present in viral DNA in several tissues with high levels of A3G expression . Our analysis CD4+ cells identified thymocytes as a cell population that has reduced A3G expression . Previous analysis of A3G expression from whole tissues did not identify thymocytes as having reduced A3G; however these results are difficult to interpret because of the lower percentage of CD4+ cells in organs other than the thymus [45] , [46] . Furthermore , the significance of A3G expression levels in the modulation of both wild-type and vif-deficient HIV replication has been previously demonstrated in Th1 and Th2 cells [47] . The implications of our findings might not be limited to HIV . Rather they might also extend to other viruses and retroelements that are restricted by APOBEC3 proteins [3] , [4] , [6]–[8] , [48] , as they may also persist as a result of reduced APOBEC3 expression that affords them the opportunity to replicate . The expansive restricting activity of the APOBEC3 family on endogenous and exogenous retroviruses serves to illustrate the broad therapeutic implications of our observations . This study also raises an important issue that must be addressed if the Vif-APOBEC3 axis is to be used to develop small molecular inhibitors of HIV replication: the well-documented ability of HIV to develop resistance to all current antiretroviral drugs . By incorporating point mutations in the relevant viral genes HIV can develop drug resistance [49] . Our observation of Vif-independent replication after direct injection into the thymus are consistent with previous work in humanized mice [50] and highlight the potential for HIV to escape the effect of a therapeutic Vif inhibitor [51]–[53] . The drug resistant virus would then be capable of systemic dissemination . However , as with other antiretrovirals , the use of combination therapy may prevent the emergence of such resistance . The thymus plays a critical role in HIV infection as it is actively involved in immune reconstitution following suppression of viremia with antiretroviral therapy . While this immune reconstitution occurs better in children than in adults , extensive thymic damage and incomplete virus suppression hinder this process [54]–[56] . Finally , it remains to be established if sublethal restriction by other innate immune defense proteins such as Tetherin , Trim-5-alpha , SamHD1 , etc . could allow the replication of other pathogenic viruses [1] . Therefore , our discovery has long lasting implications that provide an alternative view of the dynamic interplay between endogenous immune restriction factors and the broad spectrum of pathogens they control .
All animal experiments were conducted following NIH guidelines for housing and care of laboratory animals and in accordance with The University of North Carolina at Chapel Hill ( UNC-Chapel Hill ) in accordance with protocols approved by the institution's Institutional Animal Care and Use Committee . UNC-Chapel Hill protocol number 12-170 . Experiments were performed using the CCR5-tropic primary isolate HIV-1JR-CSF ( accession # M38429 ) or the CXCR4-tropic molecular clone HIV-1LAI ( accession # K02013 ) [57] , [58] . Mutations disrupting vif were made in regions that did not affect the overlapping 3′ terminus of pol or the splice acceptor site of vpr . A non-revertible 172 nucleotide deletion in the 5′ half of HIV-1JR-CSF vif ( HIVJR-CSFΔvif ) was constructed by deleting nucleotides 5138 to 5309 between the NdeI and NcoI sites . A second HIV-1JR-CSF with a potentially revertible vif ( HIVJR-CSFvifFS ) was constructed by inserting a single adenosine after nucleotide 86 in vif by site directed mutagenesis . A non-revertible 178 nucleotide deletion in the 5′ half of HIV-1LAIvif ( HIVLAIΔvif ) was constructed by deleting nucleotides 4708–4885 between the NdeI and PflMI sites [59] . All constructs were analyzed by direct DNA sequencing prior to virus production . Virus stocks were generated by transfecting proviral DNA into 293T cells using Lipofectamine 2000 ( Invitrogen ) and tissue culture infectious units ( TCIU ) were determined using TZM-bl cells essentially as we have previously reported [24] , [27] , [60] . TZM-bl Hela cells and human embryonic kidney 293T cells were cultured at 37°C , 10% CO2 in Dulbecco's Modified Eagle Medium ( Sigma ) supplemented with 10% fetal bovine serum , 50 IU penicillin , 50 µg/ml streptomycin and 2 mM L-glutamine ( Cellgro ) . CEM-SS cells were cultured at 37°C , 5%CO2 in RPMI 1640 ( Sigma ) supplemented with 10% fetal bovine serum , 50 IU penicillin , 50 µg/ml streptomycin , 2 mM L-glutamine , and 1 mM sodium pyruvate ( Cellgro ) . To generate a permissive cell line that can be infected with CCR5-tropic HIV , CEM-SS cells were transduced with the retroviral vector pBabe-CCR5 obtained from the NIH AIDS Research and Reference Reagent Program [61] , [62] . pBabe-CCR5 and the packaging vector pEQPAM were co-transfected into 293T cells using Lipofectamine 2000 ( Invitrogen ) . The culture supernatants were collected after 48 hours and filtered through a 0 . 45 µm filter . Twenty-four well plates were coated with 40 µg of Retronectin ( Takara ) and then washed with PBS+2% BSA and incubated twice with 0 . 5 mL of the vector supernatants for one hour each . CEM-SS cells ( 3×105 ) were then incubated in the vector coated wells overnight at 37°C , 5% CO2 . The following day , the vector supernatant ( 0 . 5 mL ) was added to the cells overnight . Transduced cells were selected in complete RPMI containing 0 . 5 µg/ml puromycin . Fluorescence activated cell sorting was used to isolate the CD4HiCCR5Hi population with a BD FACSAria ( Becton-Dickinson ) , collecting the top 25% . CEM-SS cells were used to propagate both wild-type and vif-deficient HIVLAI while CCR5 expressing CEM-SS cells were used for spreading infections with both wild-type and vif-deficient HIVJR-CSF . Cells ( 1×106 ) were infected with virus stocks normalized to p24Gag or tissue culture infectious units in complete RPMI containing 4 µg/ml polybrene at 37°C , 5% CO2 for 4 hours . The cells were washed extensively with PBS and cultured at 37°C , 5% CO2 in complete RPMI . Cell cultures were passaged every three days and a sample of the culture supernatant was collected for quantification of viral capsid protein by p24Gag ELISA . Human CD4+ cells from humanized mouse spleen , liver , or lung were isolated using magnetic bead sorting ( Stem Cell Technologies ) . A3G and A3F mRNA expression in cells was analyzed by quantitative RT-PCR ( qRT-PCR ) essentially as previously described [46] , [47] . Briefly , cellular RNA was extracted using the RNeasy kit ( Qiagen ) per the manufacturer's protocol including the optional treatment with RNase-free DNase ( Qiagen ) during extraction . Total RNA ( 10 ng ) was used as the template in a one-step RT-PCR reaction with the TaqMan RNA-to-Ct 1 step kit ( Applied Biosystems ) . Primers for human A3G and A3F mRNA [47] and for human TATA Box binding protein mRNA ( Applied Biosystems ) were used for amplification and human A3G and A3F mRNA levels were normalized as previously described [46] . A3G protein determination was performed by disrupting cells in lysis buffer ( 50 mM Tris , pH = 8 . 0 , 100 mM NaCl , 25 mM NaF , 25 mM benzamidine , 20 mM β-glycerophosphate , 2 mM Na3VO2 , 3 mM EDTA , 10% glycerol , and 0 . 5% IGEPAL-630 ) . Lysates were centrifuged at 13 , 000×g for 10 minutes and the supernatant fraction was prepared for SDS-PAGE gel electrophoresis . Separated proteins were transferred to nitrocellulose and immunoblotted for human A3G ( NIH AIDS reagent program #9968 ) [63] and human GAPDH ( Cell Signaling Technology #2118 ) . Protein bands were quantitated by determining density using ImageJ software ( Rasband , W . S . , ImageJ , U . S . National Institutes of Health , Bethesda , Maryland , USA , http://rsb . info . nih . gov/ij/ , 1997–2009 ) . Mice were maintained with the Division of Laboratory Animal Medicine at the University of North Carolina at Chapel Hill under specific-pathogen free conditions . Humanized mice ( BLT and NSG-hu ) were generated and analyzed for reconstitution with human hematopoietic cells including human T cells by flow cytometry essentially as previously described [20] , [21] , [24] , [25] , [27] , [28] . Humanized mice were inoculated with 3×104 or 9×104 TCIU of wild-type HIV-1LAI , 9×104 TCIU of wild-type HIV-1JR-CSF , 3 . 6×105 TCIU of HIVJR-CSFΔvif or HIVLAIΔvif , or 9×104 or 3 . 6×105 HIVJR-CSFvifFS intravenously by tail vein injection or into specific organs as indicated in the text . HIV-1 infection of humanized mice was monitored in peripheral blood by viral load analysis as previously described [27] . Tissues were harvested for evaluation of HIV-1 infection essentially as previously described [21] . Genomic DNA from mononuclear cells ( 5×105–5×106 ) from animal tissues was prepared using QIAamp DNA blood mini columns ( Qiagen ) according to the manufacture's protocol . Viral RNA was isolated from plasma using QIAamp viral RNA columns ( Qiagen ) according to the manufacture's protocol including an optional treatment with RNase-free DNase ( Qiagen ) during extraction and cDNA was generated using Superscript III Reverse Transcriptase ( Invitrogen ) . Viral DNA or cDNA was amplified by nested PCR using the Expand High Fidelity PCR System ( Roche ) . All PCR primers amplify both HIV-1JR-CSF and HIV-1LAI and were designed to anneal in regions with the fewest possible putative APOBEC3 deamination sites to avoid potential primer mismatch due to APOBEC3 induced mutagenesis . HIV regions amplified include a 1 . 5 kb region in pol ( RT: HIV-1JR-CSF 2493–4023; HIV-1LAI 2063–3595 ) , a 1 . 4 kb region including vif and vpr ( vif: HIV-1JR-CSF 4941–6399; HIV-1LAI 4511–5969 ) , and a 900 base region in the 3′ viral genome ( nef: HIV-1JR-CSF 8722–9634; HIV-1LAI 8328–9211 ) . Amplification of both RT and nef was used to assess APOBEC3 hypermutation while vif was amplified to asses APOBEC3 hypermutation and to confirm the integrity ( or restoration ) of the ORF . Primer sequences were as follows: RT outer forward primer , GCTCTATTAGATACAGGAGC; reverse primer , CCTAATGCATATTGTGAGTCTG; RT inner forward primer , GTAGGACCTACACCTGTCAAC; reverse primer , CCTGCAAAGCTAGGTGAATTGC . Vif outer forward primer , CAGGGACAACAGAGATCC; reverse primer , GTGGGTACACAGGCATGTGTGG; vif inner forward primer , CTTTGGAA AGGACCAGCAAAGC; reverse primer , GATGCACAAAATAGAGTGGTGG . Nef outer forward primer , GAATAGTGCTGTTAGCTTGC; reverse primer , CTCAAGGCAAGCTTTATTGAGG; nef inner forward primer , TAGAGCTATTCGCCACATACC; nef inner reverse , CTTTATTGAGGCTTAAGCAGTGG . Amplified viral DNA was sequenced and compared to the corresponding proviral DNA sequence used to generate the viruses using the Highlighter sequence visualization tool ( www . hiv . lanl . gov ) . One-way ANOVA with Bonferroni's multiple comparison test ( alpha level , 0 . 01 ) , Paired two-tailed t tests , and Unpaired two-tailed t tests were all performed using Prism version 4 ( Graph Pad , La Jolla , CA ) . All data were plotted as mean +/− SEM . The GenBank ( http://www . ncbi . nlm . nih . gov/nuccore ) accession numbers for HIV-1JR-CSF and HIV-1LAI are M38429 and K02013 . The GenPept ( http://www . ncbi . nlm . nih . gov/protein ) accession numbers for APOBEC3G and APOBEC3F are NP_068594 and Q8IUX4 .
|
The APOBEC3 family of proteins is a potent cellular defense mechanism capable of restricting a broad range of viruses including HIV . HIV requires a critical accessory protein , Vif , which targets APOBEC3 for degradation thereby shielding its genome from lethal mutagenesis . Previous in vitro studies have shown that in the absence of Vif , HIV can be hypermutated by APOBEC3 . This potent restrictive function of APOBEC3 has generated strong interest in developing therapeutics based on the APOBEC3/Vif axis . Here we demonstrate in vivo that CCR5-tropic HIV can be efficiently restricted by APOBEC3 . However , our results also show that CXCR4-tropic HIV can replicate independent of Vif and escape lethal restriction by APOBEC3 . Specifically , we show that thymocytes have reduced expression of A3G and A3F and that direct injection of vif-defective HIV into the thymus results in viral replication and dissemination . Despite continued Vif-independent HIV replication , the virus remained sensitive to APOBEC3 mutagenesis and was rapidly restricted in tissues with higher A3G and A3F expression . Our results provide novel insight into the restriction of HIV in vivo and identify a potentially significant defect in the innate immune defenses that protect the host cell from pathogens .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"microbial",
"mutation",
"retrovirology",
"and",
"hiv",
"immunopathogenesis",
"immunology",
"microbiology",
"host-pathogen",
"interaction",
"immunodeficiency",
"viruses",
"animal",
"models",
"of",
"infection",
"infectious",
"diseases",
"virulence",
"factors",
"and",
"mechanisms",
"viral",
"immune",
"evasion",
"hiv",
"biology",
"viral",
"persistence",
"and",
"latency",
"immunity",
"virology",
"innate",
"immunity",
"viral",
"diseases"
] |
2013
|
HIV Restriction by APOBEC3 in Humanized Mice
|
The PIK3CA gene is one of the most frequently mutated oncogenes in human cancers . It encodes p110α , the catalytic subunit of phosphatidylinositol 3-kinase alpha ( PI3Kα ) , which activates signaling cascades leading to cell proliferation , survival , and cell growth . The most frequent mutation in PIK3CA is H1047R , which results in enzymatic overactivation . Understanding how the H1047R mutation causes the enhanced activity of the protein in atomic detail is central to developing mutant-specific therapeutics for cancer . To this end , Surface Plasmon Resonance ( SPR ) experiments and Molecular Dynamics ( MD ) simulations were carried out for both wild-type ( WT ) and H1047R mutant proteins . An expanded positive charge distribution on the membrane binding regions of the mutant with respect to the WT protein is observed through MD simulations , which justifies the increased ability of the mutated protein variant to bind to membranes rich in anionic lipids in our SPR experiments . Our results further support an auto-inhibitory role of the C-terminal tail in the WT protein , which is abolished in the mutant protein due to loss of crucial intermolecular interactions . Moreover , Functional Mode Analysis reveals that the H1047R mutation alters the twisting motion of the N-lobe of the kinase domain with respect to the C-lobe and shifts the position of the conserved P-loop residues in the vicinity of the active site . These findings demonstrate the dynamical and structural differences of the two proteins in atomic detail and propose a mechanism of overactivation for the mutant protein . The results may be further utilized for the design of mutant-specific PI3Kα inhibitors that exploit the altered mutant conformation .
The PI3Kα protein is involved in cellular processes vital for cancer progression , such as cell growth , proliferation , motility , survival , and metabolism [1] . As a result , deregulation of PI3Kα signaling is one of the most frequent events leading to cancer [2] . PI3Kα uses ATP to phosphorylate the phosphatidylinositol PIP2 to PIP3 , a reaction that requires prior attachment of the enzyme to the cell membrane . Increased PI3Kα signaling may occur by several mechanisms , including somatic mutations and amplification of genes encoding key components of the PI3Kα pathway [1] . PI3Kα comprises a catalytic subunit , p110α , and a regulatory subunit , p85α . The p110α subunit consists of five domains: the adaptor-binding domain ( ABD ) , the RAS-binding domain ( RBD ) , and the C2 , helical , and kinase domains . Somatic mutations within the gene encoding p110α ( PIK3CA ) are frequently observed in a variety of human tumors , including breast , colon , endometrial cancers , and glioblastomas [3] . These mutations are scattered over the length of p110α but two hotspots account for nearly 80% of them: an H1047R substitution close to the C-terminus and a cluster of three charge-reversal mutations ( E542K , E545K , Q546K ) in the helical domain of p110α [4] . Both types of mutations can induce oncogenic transformation in cell cultures [5] , while H1047R is also able to induce tumorigenesis in transgenic mice [6] , [7] . According to structural and functional studies , these two hot spot mutations act synergistically , but independently [8]–[10] . The structure of the human [11] and mouse [8] catalytic subunit p110α has been solved by X-ray crystallography , as well as the structure of the human H1047R mutant [12] . Recent experimental data demonstrate that the H1047R mutation overactivates the enzyme by inducing dynamic changes in the kinase domain , which increase basal ATPase activity as well as expose the membrane binding regions , thereby augmenting basal membrane binding [8] , [12] , [13] . It has also been shown that the C-terminal region , where H1047R resides , is essential for catalysis . The C-terminus enhances membrane binding , while it inhibits the basal activity of the enzyme in the absence of the membrane [14]–[16] . This recent experimental work has provided mechanistic insights into the mutational activation; however , an atomic-level description of the factors that contribute to the enzyme overactivity is still missing . In the present study , we have modeled the full-length catalytic p110α subunit in the WT and H1047R mutant forms in order to gain insights into the overactivation mechanism of the commonly-expressed H1047R mutant through Molecular Dynamics ( MD ) simulations and Functional Mode Analysis ( FMA ) and have used SPR experiments to validate our results . The simulations are in excellent agreement with experimental data and allow us to provide atomic-detail insights into the mechanism of overactivation of the PIK3CA H1047R mutant by monitoring structural and dynamical elements of the WT and mutant proteins .
Five independent simulations have been performed for the full length WT and H1047R mutant p110α ( the catalytic subunit of PI3Kα ) , respectively ( Figure S1 and Table S1 ) . Cα Root Mean Square Deviation ( RMSD ) indicates that a plateau is reached in ∼100–120 ns for the WT and ∼100–135 ns for the mutant p110α ( Figure S2 ) . For the different systems , we performed Principal Component Analysis ( PCA ) using the last 50 ns of sampling for each independent simulation for the WT and H1047R mutant proteins , respectively . The 50 ns used for the PCA were chosen based on the RMSD of the Cα of the non-flexible loops ( res . number 1–7 , 231–240 , 291–330 , 410–417 , 505–530 , 863–872 , 941–952 , 1047–1068 ) . The overlap of the 2D projections of the trajectories on the first two eigenvectors indicates that the five independent simulations for the WT and the H1047R mutant proteins span the same or similar conformational phase space ( Figure S3 ) . For the analysis of the trajectories we used the last 50 ns of each trajectory , i . e . WT1 ( 100–150 ns ) , WT2 ( 100–150 ns ) , WT3 ( 100–150 ns ) , WT4 ( 120–170 ns ) , WT5 ( 100–150 ns ) for the WT and Mut1 ( 125–175 ns ) , Mut2 ( 100–150 ns ) , Mut3 ( 130–180 ns ) , Mut4 ( 135–185 ns ) , Mut5 ( 100–150 ns ) for the mutant ( Table S1 ) . It has been proposed that a second binding site , distinct from the active site , exists in PI3Kα as demonstrated in a recent WT crystal structure [8] . Representative conformations from the first three cluster representatives of the WT protein ( taken from simulation WT1 , see Table S2 ) were submitted to Q-SiteFinder server [17] for ligand binding site prediction in the kinase domain . As seen in Figure 1 , the existence of a new binding pocket , distinct from the active site and close to H1047 in the WT protein , is confirmed . To determine key features of the mutant protein that may result to overactivity with respect to the WT , the structure and dynamics of interaction network in both proteins were monitored . Functionally important structural elements of the PI3Kα kinase domain are highlighted in Figure 2 . It has been suggested that intermolecular interactions between helix kα11 ( res . numbers 1031–1047 ) , which precedes the C-terminal tail , and the catalytic loop ( res . numbers 909–920 ) shield the conserved catalytic DRH motif from performing futile ATP hydrolysis by His-917 [15] , [16] . Indeed , our calculations show that an interaction network , tightly coupled to His-1047 , accurately controls the DRH motif and retracts it from the vicinity of the active site: the hydrogen bond between the WT His-1047 backbone nitrogen and Met-1043 backbone oxygen as well as the hydrogen bond between the WT His-1047 π-imidazole ring nitrogen ( see Figure S4 for definition ) and Met-1043 backbone oxygen stabilize the last α-helix of the protein , kα11 ( see Figure 3 A , hydrogen bonds have frequencies 46±7% and 64±5% , respectively ) . Statistical errors have been calculated as standard deviations from five independent simulations . The autocorrelation functions of the hydrogen bonds time series have been calculated and found to converge within the simulation time . kα11 is further stabilized by a frequent hydrogen bond between the WT Gly-1049 nitrogen and Asn-1044 backbone oxygen ( 72±8% ) . As a result , a hydrogen bond between the His-1047 τ-imidazole ring nitrogen and the amide hydrogen of activation loop residue Leu-956 occurs at high frequency ( 92±1% ) in four out of five simulations of the WT p110α , in agreement with experimental data ( Table S3 ) [12] . In Simulation WT1 , this hydrogen bond is infrequent ( 1% ) , however , Leu-956 backbone oxygen is frequently hydrogen bonded with Thr-1053 backbone oxygen ( 90% ) , leading to a similar conformation of the WT protein kα11 helix in all five simulations ( Figure S5 ) . As His-1047 is kept tightly controlled in the WT , Arg-949 forms frequent hydrogen bonds with Asp-915 of the conserved DRH motif in four out of five of WT simulations ( 78±6% , Table S3 ) . In the remaining simulation ( Simulation WT3 ) , Arg949 side chain nitrogen atoms form hydrogen bonds with Asp-939 backbone oxygen with a frequency of 78% . In all five simulations , we consistently observed a hydrogen bond formed between Arg-916 of the DRH motif and Asp-933 of the DFG motif in agreement with Miller et al . [15] . In stark contrast , this extensive hydrogen bond network observed in the vicinity of the WT His-1047 is not present in the mutant structure . Arg-1047 breaks the hydrogen bond with Leu-956 . Consequently , the His-to-Arg substitution at position 1047 destabilizes the kα11 α-helix by disruption of the hydrogen bonds between Arg-1047 and Met-1043 as well as Gly-1049 and Asn-1044 ( Figure 3 B , Table S3 ) . The disruption of the α-helix allows Asp-1045 to hydrogen bond with Arg-949 in three out of five mutant simulations . In the other two simulations , the final residues of kα11 , Asp-1045 , Ala-1046 , and Arg-1047 , are in an extended conformation and do not form any hydrogen bonds with the activation loop . In all five simulations of the H1047R mutant the activation loop residue Asp-951 forms frequent hydrogen bonds with either residues Gly-912 , Phe-909 or Asp-939 from the activation loop . In turn , the hydrogen bond between Arg-949 and Asp-915 is abolished in the mutant H1047R p110α . This destabilization of Asp-915 allows the side-chain of His-917 of the DRH motif ( which in the active PI3Kα conformation participates in the ATP hydrolysis ) to point towards the active site ( Figure S6 ) . This conformation of the His-917 was also observed in Ref . 17 , where His-745 of PI3Kγ in its active form points towards the active site as compared to the catalytic His-807 of Vsp34 ( the primordial PI3Kα ) that points away from the ATP site ( see Figures 2 C and 3 of Ref . [15] ) . Similarly to the WT simulations , the hydrogen bond between Arg-916 of the DRH motif and Asp-933 of the DFG motif is also evident , though to a lesser extent ( WT frequency , 79±6%; mutant , 50±14% ) . To further probe intermolecular interactions that govern the observed differences between WT and mutant , we have calculated the average distances of the side chains of the WIF motif residues ( res . numbers 1057 , 1058 , and 1059 ) . This motif is conserved in class I and class II PI3Ks as a triplet of hydrophobic residues and has been proposed to be crucial for lipid binding in the C-terminal region [8] . Stacking of the hydrophobic side chains of the WIF motif is evident in our simulations as has been suggested by experiments and may play a role in lipid binding ( Figures S7 , S8 and Table S4 ) [8] , [18] . In particular , we observe stacking between Ile-1058 and Phe-1059 in the largest part of the trajectory of both the mutant and WT simulations , while Trp-1057 and Ile-1058 are occasionally found within stacking distance in both proteins . We also monitored the effect of the H1047R mutation on the solvent accessibility of kα11 and kα12 ( C-terminal tail , residues 1032–1068 ) . Our results demonstrate that residues 1032–1047 ( kα11 ) are significantly more solvent exposed in the H0147R mutant , while residues 1048–1069 ( kα12 ) are more solvent exposed in the WT ( Table S5 ) . In particular , residues Met-1043 , Trp-1057 as well as the mutated residue 1047 have higher solvent accessible area in the mutant , whereas Met-1055 and Asp-1056 are more solvent accessible in the WT protein ( Figure S9 ) . The examination of the polar contact network of the WT and the mutant proteins ( i . e . salt bridges and hydrogen bonds ) indicates differences between the two protein forms within the active site ( Figures 4 , S10 , and S11 ) . The hydrogen bond between Asp-810 of the affinity pocket and Phe-934 of the activation loop ( DFG motif ) occurs at a frequency of 92±1% of the simulation time in the WT , while it has zero or low occurrence in three out of five simulations of the mutant . In the other two simulations of the mutant the Asp-810-Phe-934 hydrogen bond has an average frequency of 85±3% , this , however , does not affect the conformation of the activation loop which is similar in all five simulations of the mutant ( Figure S11 ) . Met-772 is hydrogen bonded through its backbone nitrogen to the backbone oxygen of Pro-778 in 75±12% of the simulation time in the WT and in 89±1% of the mutant trajectories . Loop residues 771–780 interact with the phosphates of the ATP and are known to comprise the P-loop ( see Table 1 ) [18] . Despite this common interaction within the P-loop residues , two hydrogen bonds between the backbone oxygen and nitrogen of both Arg-770 and Trp-780 occur with higher frequency in the mutant protein than in the WT , 73±9% and 87±1% for the mutant and 37±8% and 39±8% in the WT , leading to a more compact conformation of the P-loop in the case of H1047R mutant ( Figure 4 B ) . This closed conformation is further enforced through a hydrogen bond between the backbone oxygen of Ser-774 and the backbone nitrogen of Arg-777 in the mutant simulations ( 45±7% ) . The same hydrogen bond is either infrequent or completely absent in the WT simulations , allowing for a more wide conformation of the P-loop . Estimation of the solvent exposed surface area in the active site of the WT and H1047R p110α subunit showed that both proteins have similar solvent accessibility ( Figure S12 , Table S5 ) . This finding is in agreement with previous observations [19] . PI3Kα attaches to the cell membrane in order to retrieve the PIP2 substrate and transform it to PIP3 . It has been reported that the H1047R mutation augments the interaction between PI3Kα and the membrane; the enzymatic activity of PI3Kα H1047R is increased compared to the WT protein upon interaction with phosphatidylserine ( PS ) and cancer liposomes [12] . Further lipid-PI3Kα association studies [8] , using the WT and H1047R mutant proteins and neutral and anionic PS liposomes , showed that lipid binding of the mutant protein is many fold higher than that of the WT enzyme . The inner leaflet of the cell membrane has a net negative charge resulting from the predominance of phosphatidylserine and phosphatidylinositol on the cytosolic face of the plasma membrane . Thus , to rationalize the fact that H1047R PI3Kα binds to cell membranes with higher affinity , we performed electrostatic potential calculations on the proposed membrane binding domains , which involve residues 863–873 , 721–727 , the end of the C-terminal tail , and residues along the activation loop [11] , [12] . The calculated electrostatic potential on the surface of the two proteins reveals major differences in the positive charge distribution of membrane binding regions ( Figures 5 and S13 ) . In the H1047R p110α , the C-terminus protrudes to the plane of the membrane as a hydrophobic tail , while it is surrounded by enhanced positive charge accumulated along the activation loop , the kα7/kα8 ( 966–974 ) , and kα6/kα5 ( 863–873 ) elbows ( Figures 5 B and S13 B , D ) . These regions of the kinase C-lobe , which are pronouncedly less positively charged in the WT ( Figure 5 A and S13 A , B ) , have been found to interact with neutral and anionic membranes only in the case of H1047R mutant [13] . Interestingly , the mutant also exhibits higher positive charge than the WT along the other membrane binding site ( residues 721–727 on the N-lobe ) which was protected in both WT and H1047R HDX-MS experiments by PIP2 phospholipid vesicles [13] . Positive charge is also detected on loops 343–351 and 410–418 of the mutant C2 domain , which have also been proposed to contact the cell membrane [11] . In order to validate our MD results and compare lipid binding of WT and mutant PI3Kα , we employed Surface Plasmon Resonance ( SPR ) to monitor direct binding to cancer liposomes . Liposomes were prepared from total lipid extractions from HCT116 human colorectal cancer cells and were enriched with 2% PIP2 . As cell membranes contain mostly negatively charged lipids , we expect the liposomal surface to be predominantly negatively charged . In titration experiments , a concentration dependent binding of both WT ( Figure 6 A ) and mutated PI3Kα ( Figure 6 B ) to liposome surfaces was observed . H1047R bound liposomal membranes at repeatedly higher levels than equimolar WT protein ( typically ∼2–2 . 5 fold at 40 nM at the peak of association which corresponds to end of injection; Figure 6 B ) . This increase in lipid binding characterizes the behavior of the inactive forms of the enzyme as no activating RTK phosphopeptide was included in the experiments . No significant binding was observed to control surfaces for both proteins , demonstrating PI3K specificity for lipids . In the absence of PIP2 , we could not detect substantial PI3Kα-liposome interaction , indicating that our results do not depend on endogenous PIP2 in cancer liposomes . Comparisons were performed using data obtained from the same injection and liposomal surface , in a “one shot” grid approach ( see Materials and Methods section ) . Root mean square fluctuation ( RMSF ) analysis of the MD trajectory reveals that three regions important for enzyme function exhibit different mobility in the WT and the mutant ( Table S6 ) . The activation loop ( 933–958 ) has an average RMSF of 1 . 26±0 . 12 Å in the WT and 1 . 66±0 . 16 Å in the mutant . Moreover , the catalytic loop is more flexible in the mutant with an RMSF of 1 . 04±0 . 05 Å compared to an RMSF of 0 . 81±0 . 04 Å in the WT , as well as the P-loop . The C-terminus ( residues 1048–1068 ) sporadically forms helical turns ( see Videos S1 and S2 ) in both the mutant and the WT proteins , but its overall flexibility remains high . The RMSF is 3 . 40±0 . 90 Å in the WT and 3 . 42±0 . 36 Å in the mutant; however their conformation is entirely different . Although the initial configurations of the C-termini of mutant and WT proteins were modeled to occupy the same area in space ( Figure S14 shows the kα7/kα8 and kα6/kα5 elbows ( blue ) to be aligned very well in their initial conformation ) , during the course of the simulation the WT C-terminus always shields the ATP binding pocket , whereas the C-terminus of the mutant is pulled above the ATP site ( see Videos S1 and S2 ) . Kinases are known to exhibit two characteristic large-scale motions in the absence of ATP: a bending motion centered at the hinge region , between the N- and the C-lobe , and a twisting motion of the N- lobe with respect to the C-lobe [20] . Although these motions can be described by Principal Component Analysis ( PCA ) , they are not captured entirely by a single principal mode . Thereby , Functional Mode Analysis ( FMA ) was implemented in order to identify collective motions related to the hinge bending motion and the C- and N- lobe twisting motions . The functional quantity that yielded the highest correlation to the hinge bending motion was the distance between the Cα carbons of Leu-781 and Met922 of the active site ( dLM ) . Residues Leu-781 and Met-922 were selected to quantify the hinge bending motion as they lie on opposite sites of the catalytic cleft and their distance is directly related to the opening and closing of the active site . In contrast to the hinge bending motion , which occurs in two dimensions ( linear ) , the twisting motion occurs in three dimensions . For this motion , the highest correlated functional quantity was the RMSD of the Cα of the active site residues ( RMSDact ) , which is a non-linear metric . For dLM , the collective vector α was optimized by maximizing the Pearson's correlation coefficient ( R ) , yielding linear models for the WT and mutant dLM . We used the first 35 ns of the production run for model building and the rest 15 ns for cross-validation ( Figure S15 C–H ) . To avoid over-fitting of the model in the selection of the basis set , the Pearson's correlation coefficients of the model-building ( Rm ) and the cross-validation set ( Rc ) were plotted as a function of the number of eigenvectors used as a basis set ( Figures S15 A and S15 B ) . The hinge bending motion between the two lobes of the kinase domain from the WT and H1047R p110α is illustrated in Figures S15A , B and video S3 . Our analysis shows that the P-loop is closer to the catalytic loop in the WT than in the mutant throughout the course of the motion ( video S3 ) . For the description of the twisting motion , the RMSDact was optimized by maximizing the mutual information ( MI ) coefficient ( see Supporting Information Text S1 , section A6 for more details ) . The MI is used to quantify non-linear , higher order correlation . We used the first 40 ns of the production phase for model building the last 10 ns for cross-validation ( Figures S16 A–D ) . For the optimization of the non-linear model with the MI , we used less than 20 eigenvectors to avoid over-fitting . As shown in Figures S16 A , B , the difference between Rm and Rc reaches a minimum when the number of used eigenvectors is 17 in the WT and 13 in the mutant . The two basis sets yielded a Pearson's correlation values of 0 . 86 and 0 . 87 for the WT and the mutant trajectory , respectively ( Figures 7 A , B ) , which denote high correlation between the RMSDact and the twisting motion of the kinase lobes . In both the mutant and the WT , the P-loop lies on the same plane , however , as the motion progresses , it is shifted outwards in the mutant with respect to the WT , broadening the solvent accessibility of the mutant catalytic cleft ( Figures 7 C , D and video S4 ) . A more open conformation of the P-loop in the mutant structure was also observed through our polar contact analysis within the active site of the WT and mutant proteins ( see above ) . A greater catalytic cleft may lead to enhanced substrate accessibility . Moreover , the catalytic loop of the WT comes closer to the ATP binding site thus reducing the volume of the pocket . The mutant activation loop lies below the activation loop of the WT in the starting position , but this positioning is reversed in the final position . The average conformations of the kinase domains show that the P-loop of the WT p110α curls inwards , towards the ATP-binding cavity when compared with the H1047R p110α , which results to a greater catalytic cleft in the mutant protein ( Figure S17 ) . Moreover , in order to quantify the overlap between the mutant and WT trajectory eigenspaces , we calculated the Root Mean Square Inner Product ( RMSIP ) for all corresponding eigenvectors arising from PCA for the kinase domain ( see Text S1 , section A5 and Table S7 ) . The mutant and the WT trajectory RMSIP yielded a normalized value of 0 . 23±0 . 01 for the kinase domain PCA and for the five independent trajectories , indicating that the eigenspaces of the WT and the mutant are different . In comparison , the RMSIP of the WT trajectories is 0 . 38±0 . 03 and for the mutant RMSIP = 0 . 42±0 . 01 . In other words , the motions along each PC of the WT did not correspond to the motions of the equivalent principal component of the mutant as shown by their overlap , which was 23% .
Results presented herein lead to a model of the overactivation mechanism of the commonly-expressed PIK3CA mutant H1047R based on structural and dynamic differences with its WT counterpart , schematized in Figure 8 . SPR experiments show that the H1047R mutant binds liposomal membranes with higher competence . This finding is rationalized through MD simulations and subsequent electrostatic potential calculations , which verify that the mutant protein accumulates positive charge on the membrane binding domains of PIK3CA . This accumulation of positive charge explains the experimental finding that the mutant binds membranes rich in anionic lipids with higher capacity than the WT . Previous studies have shown that the C-terminal tail of the mutant is more solvent exposed than its WT counterpart , which is also confirmed through our MD simulations . Moreover , we verify the prediction of a second , unexpected binding pocket close to the area of the mutation , recently discovered by X-ray crystallography . Following the agreement with experimental results , we extend our studies to highlight the series of events that lead to the overactivation of this protein kinase mutant . Our results support an auto-inhibitory role of the C-terminal tail in the WT protein , which strictly controls the DRH motif to limit its access to the catalytic site . We propose that the weakening of this role in the H1047R mutant through loss of crucial intermolecular interactions is a plausible explanation of the elevated kinase activity of the enzyme . One major difference between the polar contact network of the WT and H1047R is the loss of the hydrogen bond connecting Arg-949 and catalytic Asp-915 of the DRH motif in the mutant , which occurs in the WT in 78±6% occurrence , while it is absent in the mutant . Arg-949 is known for conferring specificity to PIP2 in PI3Kα [8] . The same functional role has also been reported in the γ isoform [21] . The difference in the polar contacts , position , and consequently the availability of the positively charged residue , Arg-949 ( along with Lys-948 and Arg-951 ) , alters the configuration of the activation loop and exposes its positive charges , making it seemingly more capable to bind to the membrane and accommodate negatively charged phosphoinositide headgroups . We also suggest that the abrogation of the Arg-949 - Asp-915 interaction in the mutant may contribute to the overactivity of the enzyme , since when these residues are left unrestrained they have enhanced access to the catalytic site [8] . In the H1047R p110α , Arg-1047 disrupts the last helical turn of kα11 and unwinds the hairpin observed in the crystal structure ( compare Figures 3 A and 3 B ) [12] . One plausible explanation is that substitution of the bulky imidazole ring by the longer aliphatic straight chain capped with the positively charged guanidium group , rules out the hydrogen bond interaction with the Leu-956 backbone amine , which is conserved throughout the trajectory of the WT , but is completely absent in the mutant ( Figures 2 , S5 and S6 , Table S3 ) . Mandelker et al . [12] also accentuates the loss of this hydrogen bond in the crystal structures of H1047R p110α/p85α-niSH2 complex ( PDB codes 3HIZ , 3HHM ) and suggests that it is this interaction that stabilizes the WT activation loop . According to our data , the activation loop in the mutant structure is more flexible than in the WT and this can partially explain the oncogenic phenotype of the H1047R p110α ( Table S6 ) . On the other hand , the loss of interaction between residues 1047 and 956 , evoke the disruption of the simultaneously occurring hydrogen bond between Leu-956 backbone carbonyl and Thr-1053 side chain carboxyl . These two stable hydrogen bonds , mediated by Leu-956 in the WT , preserve the hairpin formation and hold the C-terminal tail in an arrangement that favors interaction with the DRH motif of the catalytic loop . This interaction with the DRH motif in the WT keeps His-917 , which participates in ATP hydrolysis , pointing away from the active site thus preventing ATP hydrolysis more efficiently . This conformation of His-917 has also been observed in the inactive form of Vsp34 ( the primordial PI3Kα ) , where the catalytic His-807 points away from the ATP site ( see Figures 2 C and 3 of Ref . [15] ) . In contrast , in the mutant PI3Kα structure , His917 points towards the active site , a conformation that is also observed in the structure of the active PI3Kγ , where His-745 is the residue participating in catalysis [15] . Additionally , the C-terminal tail of the H1047R mutant is significantly more solvent exposed in the H1047R mutant compared to the WT . It is worth noting that Hon et al . [8] reported that all mutations in the kα11/kα12 elbow ( H1047L , H1047R , G1049R ) exhibited a few fold higher levels of hydrophobic binding to neutral lipids and electrostatic binding to negatively charged lipids than the WT p110α , which may be associated with enhanced solvent accessibility of the mutant kα12 region and in particular higher solvent accessible area per residue , as indicated by Figure S9 . The importance of the final C-terminal helices of the p110α catalytic subunit in the regulation of the enzyme has been previously highlighted [14]–[16] . This regulatory arch encircles the catalytic and activation loops and is believed to control the enzymatic activity [14] . The last helix , kα12 , which is disordered in p110α and p110δ , has two additional roles: ( a ) an activating role when in contact with the membrane [8] , ( b ) an auto-inhibitory role when the enzyme is not interacting with the membrane [14]–[16] . In the latter case , kα12 locks the catalytic loop in an inactive state , presumably by shielding the conserved catalytic DRH motif ( 915–917 ) from performing futile ATP hydrolysis . This second role has been inferred from inspection of the crystal structures of PI3K isoforms α , β , γ , δ and their paralogue in Drosophila melanogaster Vps34 , as well as from truncation of the C-terminus of p110α , p110β and Vps34 that resulted to enhanced basal ATPase activity in the absence of lipid substrate [8] , [15] , [16] . These results verify the self-inhibitory role of the C-terminus . Remarkably , the C-terminus interaction with the activation loop is relieved in the H1047R p110α ( Tables S8 , S9 , S10 ) , which may well be part of the explanation of the enhanced kinase activity of that enzyme . The highly conserved DFG motif ( 933–935 ) at the beginning of the activation loop is believed to adopt different configurations during the various steps of the catalytic cycle of kinases [22] In PI3K , the aspartate side chain of DFG ( Asp-933 ) bends in order to form polar contacts with the last phosphate group of ATP [23] . This is believed to be the “in” conformation which designates the active state , whereas in the “out” conformation the aspartate side chain extends straight towards the ribose ring . Our results show that both WT and mutant Asp-933 adopt an “in-like” conformation , albeit Asp-933 of the WT structure is frequently turned away from the ATP binding site due to a high-frequency hydrogen bond with Arg-916 and a low frequency hydrogen bond with the Gly-935 backbone amine ( Table S3 and videos S1 and S2 ) . Both these polar contacts have significantly lower frequencies in the H1047R p110α and thus the mutant Asp-933 shows a tendency to assume an “in-like” conformation more frequently than the WT , providing an advantage to the former regarding ATP binding . Moreover , Asp-933 in both protein structures forms a salt- bridge with residue Lys-802 of the active site . The equivalent of residue Lys-802 in PI3Kγ crystal structure ( Lys-833 ) forms a hydrogen bond with the pan-PI3K inhibitor PIK-90 [24] . It is , therefore , plausible that Lys-802 plays a crucial role in the design of inhibitors like PIK-90 . Lastly , in accordance with a previous simulation [19] , the change in the orientation of Asp-933 is accompanied by a flip of the Phe-934 side chain that renders it more exposed to substrates entering the ATP-binding site . These observations may provide an important basis for the design of mutant selective inhibitors . Changes in the polar contact network within the active site were also observed . We discern a change in the hydrogen bonding frequency between the side chain of Asp-810 and the backbone of Phe-934 from the DFG motif . This hydrogen bond loss could be exploited in the design of mutant-specific inhibitors targeting Asp-810 , given that ligand binding in the WT active site would have to overcome the additional enthalpic cost for breaking these two frequent hydrogen bonds . Moreover , the specificity pocket residue Trp-780 is involved in two hydrogen bonds with Arg-770 with much higher frequency in the mutant than in the WT protein , providing valuable information for the design of inhibitors targeting the H1047R p110α . Finally , through FMA we show that differences in the twisting motion of the kinase lobes exist , with the mutant having a greater opening of the catalytic cleft , which may favor ATP binding and thus influence kinase activity . Changes in the twisting motion have been previously proposed to alter kinase activity [20] . Through the FMA and polar contact analysis , we observe a wider conformation of the P-loop relative to the ATP pocket in the mutant structure , while at the same time it is more compact compared to the WT structure , which could lead to enhanced accessibility of the active site . Overall , understanding how the H1047R mutation causes the enhanced activity of the protein in atomic detail is of paramount importance for developing mutant-specific therapeutics for cancer .
A full description of the methods can be found in the Text S1 ( sections A1–A6 ) . Briefly , two models of the WT p110α were constructed: in Model 1a the missing C2 domain loop residues 415–423 were created through loop modeling ( Figure S18 B , Text S1 , section A2 ) . In Model 1b , residues 335–361 and 401–428 were re- constructed through homology modeling ( Text S1 , section A1 and Figure S18 C ) , using as a template the solution NMR structure of the human C2 domain with PDB accession code 2ENQ ( Figure S18 D ) , due to the low electron density of the WT p110α structure ( 2RD0 ) at this area ( Figure S19 ) . The rest of the missing loops of 2RD0 were created through loop modeling and share the same coordinates in both Models 1a and 1b . Model 2 was built from the PDB structure 3HIZ [12] . C2 domain residues 335–361 and 402–428 were rebuilt from structure 2ENQ and residues 857–884 were rebuilt using PDB structure 3HHM as a template through homology modeling due to low electron density in this region ( Text S1 , section A3 ) . The MD trajectories for the WT and H1047R p110α were generated with NAMD v2 . 7 [25] , using the CHARMM22 all-atom force field with the CMAP correction [26] , [27] and the TIP3P water model [28] . Five independent simulations for each protein were performed . Moreover , for the WT protein , we simulated both Model 1a and one of Model 1b in order to assess the effect of the remodeled C2 domain . All proteins were solvated into a cubic box large enough to ensure a 30 Å minimum separation of the protein from its periodic images . Na+ counter-ions were placed randomly in the system to neutralize the total charge ( see Text S1 , section A4 for more details ) . The production runs were performed under constant pressure , temperature , and number of particles ( NPT ) . The convergence of our simulations was evaluated using the total Cα carbon RMSD , while to ensure that each set of independent simulations corresponds to the same conformational protein phase space , we plotted the 2 d projection of the trajectories on the first two eigenvectors of each simulation ( Text S1 , section A5 ) . The trajectory was analyzed with nMOLDYNv3 . 0 . 8 [29] , MMTK-2 . 7 . 4 [30] , GROMACS tools v4 . 5 . 5 [31] , NAMD v2 . 7 [25] , [32] , PDB2PQR [32] , APBS [33] , and VMD [34] . Functional Mode Analysis ( FMA ) was performed as discussed in Ref . [35] . Binding site prediction was performed with the QSiteFinder web server [17] . For more details on trajectory analysis see the Text S1 , section A5 . The ProteOn biosensor ( Bio-rad ) was used for SPR analysis . ProteOn uses a unique 6×6 chip array ( positions L1 to L6 are vertical and positions A1 to A6 are horizontal ) that allows running experiments in a grid format . The LCP memLayer kit ( Bio-Rad ) was used to tether liposomes ( in two consecutive layers ) on the chip up to ∼3500 RU , using the vertical ( L ) channel direction . Blank control channels ( L direction ) that were treated equivalently to liposome loaded channels but lacked liposomes , were used for data reference . Appropriate concentrations of WT and H1047R PI3K ( diluted in SPR running buffer just before run ) were injected over the chip using the parallel ( A ) channels . To ensure high experimental uniformity and confidence in SPR data comparisons , each binding experiment was performed in “one shot” , i . e . both WT and H1047R were injected in the same parallel injection ( occupying different A positions ) . Background PI3K binding and bulk effects within each injection were referenced using blank L channels . Experiments were repeated at least 3 times , using fresh liposomes loadings . SPR running buffer: 10 mM NaPi pH 7 . 4 , 150 mM NaCl , 0 . 1 mg/ml BSA . Pure human WT p110α/p85α and mutant p110α ( H1047R ) /p85α heterodimers were purchased from Millipore . Protein activity was verified using PI3-Kinase HTRF Assay ( Millipore ) . Proteins in 50 mM Tris/HCl pH 8 . 0 , 300 mM NaCl , 0 . 1 mM EGTA , 0 . 03% Brij-35 , 270 mM sucrose , 0 . 2 mM PMSF , 1 mM benzamidine , 0 . 1% 2-mercaptoethanol were aliquoted and stored at −80°C before use . Liposomes have been prepared using lipids isolated from the cancer cell line HCT116 , which carries a mutation in exon 20 of PIK3CA ( H1047R ) [2] , according to Folch's method [36] . Then , L-α-phosphatidylinositol-4 , 5-bisphosphate ( brain , porcine , ammonium salt , Avanti Polar Lipids , Inc . ) was added to the extracted lipids ( PIP2 concentration was 2% of the total lipids ) , and the mixture was immediately dried under N2 stream . When the sample was completely dried , lipids were left for another 30 minutes under the N2 stream , followed by speed-vacuum for 1 hr . Subsequently , water was added to the dried lipids and the mixture was incubated at room temperature for 1 hr , while it was vortexed every 10 minutes . The liposomal preparation was subjected to 5 freeze/thaw cycles and was sonicated in a waterbath for 30 minutes . Finally , the liposomes were extruded using the Avanti mini-Extruder apparatus , according to manufacturer's instructions , in order to obtain a homogeneous preparation of unilamelar liposomal vesicles at a size of 100 nm .
|
The PI3Kα protein is involved in cellular processes such as cell growth , division , and formation of new blood vessels ( angiogenesis ) that aid cancer cell survival . In certain types of cancer cells , PI3Kα is found to be altered compared to healthy cells . These PI3Kα alterations , called mutations , are found in 27% of breast cancer patients , 24% of endometrial cancer patients , and 15% of colon cancer patients . PI3Kα mutations cause the protein to become overactivated and may contribute to tumor growth . The most common PI3Kα amino acid mutation is a histidine changed to an arginine: H1047R . Understanding how the H1047R mutation overactivates PI3Kα is central to developing therapeutics for cancer patients who bear PI3Kα mutations . To this end , we performed simulations and experiments of the mutated and physiological proteins to explain why the mutant protein becomes overactivated . Our results indicate structural and dynamical differences between the mutant and physiological proteins that may affect the PI3Kα function . Based on these differences , we propose a mechanism that highlights the series of events that lead to the mutant H1047R PI3Kα overactivation . This study provides insights into developing mutant-specific PI3Kα inhibitors that exploit the altered conformation of the mutant with respect to the physiological protein .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"physics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"biophysics",
"biophysical",
"simulations"
] |
2014
|
Investigating the Structure and Dynamics of the PIK3CA Wild-Type and H1047R Oncogenic Mutant
|
Rer1 is a retrieval receptor for endoplasmic reticulum ( ER ) retention of various ER membrane proteins and unassembled or immature components of membrane protein complexes . However , its physiological functions during mammalian development remain unclear . This study aimed to investigate the role of Rer1-mediated quality control system in mammalian development . We show that Rer1 is required for the sufficient cell surface expression and activity of γ-secretase complex , which modulates Notch signaling during mouse cerebral cortex development . When Rer1 was depleted in the mouse cerebral cortex , the number of neural stem cells decreased significantly , and malformation of the cerebral cortex was observed . Rer1 loss reduced γ-secretase activity and downregulated Notch signaling in the developing cerebral cortex . In Rer1-deficient cells , a subpopulation of γ-secretase complexes and components was transported to and degraded in lysosomes , thereby significantly reducing the amount of γ-secretase complex on the cell surface . These results suggest that Rer1 maintains Notch signaling by maintaining sufficient expression of the γ-secretase complex on the cell surface and regulating neural stem cell maintenance during cerebral cortex development .
Newly synthesized proteins are primarily assessed by the ER quality control system for correct folding , modification , and complex formation . Nevertheless , some immature or misfolded proteins escape from the ER to post-ER compartments . However , they are recognized by several retrieval receptors at the early-Golgi compartment and are hence returned to the ER . The latter mechanism is referred to as the “early-Golgi quality control system” . Rer1 is a well-conserved early-Golgi membrane protein in eukaryotes , which plays important roles in the early-Golgi quality control system . Rer1p was originally identified in yeast as a receptor for the retrieval of ER membrane proteins from the early-Golgi to the ER via the COPI-mediated retrograde transport pathway [1–4] . Recently , the roles of Rer1 have been extended to the ER retention of immature or misfolded mutant membrane proteins related to some ER misfolding diseases [5 , 6] . Importantly , Rer1 also modulates the formation of multimeric membrane protein complexes by retaining unassembled components in the ER . In yeast , Rer1p recognizes an unassembled iron-transporter subunit Fet3p and retains it in the ER until Fet3p forms an appropriate complex with its partner , Ftr1p [7] . These observations suggest that Rer1 functions as a sorting chaperone to support the formation of multimeric membrane protein complexes [4 , 7] . γ-Secretase is a multimeric membrane protein complex , which comprises presenilin-1 ( PS1 ) , nicastrin ( NCT ) , anterior pharynx-1 ( APH-1 ) , and presenilin enhancer γ-secretase subunit ( also known as PEN2 ) [8–13] . Unassembled subunits of the γ-secretase complex are retained or degraded in the ER [14] . Only properly assembled γ-secretase is transported to post-Golgi compartments and functions as an intramembrane aspartyl protease , which cleaves type I transmembrane proteins such as amyloid precursor protein ( APP ) and Notch [15–17] . In this process , Rer1 binds unassembled NCT and/or PEN2 and retains them in the ER [18 , 19] . Since siRNA-mediated knockdown of Rer1 enhances complex formation and the activity of the γ-secretase complex in cultured cells , Rer1 has been considered to negatively regulate γ-secretase function via competition with APH-1 for NCT binding [19–21] . The present study aimed to investigate the role of the Rer1-mediated quality control system in mammalian development . First , we generated Rer1 knockout mice and determined whether it is essential in early mouse development . Furthermore , we examined the role of Rer1 in mouse cerebral cortex development . Our study hence reveals the primary role of Rer1 in Notch signaling in the cerebral cortex development and in the regulation of the neural stem cell population .
To examine the physiological roles of Rer1 in mammalian development , we generated a knockout mouse using an embryonic stem ( ES ) cell line ( EUCOMM ) , in which a reversible gene-trap cassette , FlipRosaβgeo , is inserted in the intron 1 of Rer1 ( S1A and S1B Fig ) . We first generated Rer1 heterozygous gene-trap ( Rer1+/trap ) mice , using this ES cell line . We confirmed a single gene-trap cassette insertion into the Rer1 locus in Rer1+/trap mice via Southern blot analysis ( S1A and S1C Fig ) . Although the Rer1+/trap mice showed normal gross morphology and fertility , these mice were 10–20% lighter than Rer1+/+ mice ( S1A and S1B Fig ) . The protein level of Rer1 was reduced in Rer1+/trap mice ( S2C Fig ) , suggesting that heterozygous loss of Rer1 results in haploinsufficiency in body size ( S2A–S2C Fig ) . Furthermore , we attempted to generate Rer1-homozygous gene-trap mice ( hereafter Rer1trap/trap ) by intercrossing Rer1+/trap mice . However , Rer1trap/trap mice were embryonic lethal , as reported previously [22] , indicating that Rer1 plays an essential role in mouse early development ( S1D Fig ) . To circumvent the developmental lethality of Rer1 deficiency , we crossed Rer1+/trap mice with CAG-NLS-FLPe transgenic mice ( S3A–S3C Fig ) . Rer1inv-trap homozygous mice were born at the predicted Mendelian ratio , indicating that the embryonic lethality by Rer1 gene inactivation using the gene trap cassette was primarily canceled by flipping it to the noncoding strand in vivo . Since Rer1 is involved in the formation of the γ-secretase complex , we focused on the roles of Rer1 in the mouse brain development . To inhibit Rer1 expression in the developing cerebral cortex by reinverting the inverted trap cassette in the forebrain of Rer1inv-trap homozygous mice , we crossed Rer1inv-trap homozygous mice with Emx1-Cre transgenic mice , which express a Cre recombinase only in the forebrain of embryos and the cerebral cortex of adult mice [23] . The resulting Rer1+/inv-trap; Emx-Cre mice were crossed with Rer1inv-trap homozygous mice to generate forebrain-specific Rer1 conditional knockout mice ( Rer1inv-trap/ inv-trap; Emx-Cre mice ) . The forebrain-specific Rer1 conditional knockout ( Rer1 cKO ) mice were born at the Mendelian ratio . However , half of forebrain-specific Rer1 cKO mice died soon postpartum . The remaining forebrain-specific Rer1 cKO mice survived for more than one year . The forebrain-specific Rer1 cKO mice showed limb-clasping reflexes when suspended by their tails , whereas control mice ( Rer1+/inv-trap , Rer1inv-trap /inv-trap , Rer1+/inv-trap; Emx-Cre ) extended their limbs ( Fig 1A ) . In the rotarod test to assess motor coordination , control and forebrain-specific Rer1 cKO mice showed a similar behavior on days 1 and 2 , suggesting their normal motor function and learning ( Fig 1B ) . We also examined anxiety and general locomotor activity levels of Rer1 cKO mice via the open field test ( Fig 1C ) . Interestingly , forebrain-specific Rer1 cKO mice spent more time than control mice around the center of the field , suggesting that Rer1 loss in the forebrain reduces anxiety ( Fig 1D ) . However , total distance traveled by control and forebrain-specific Rer1 cKO mice were comparable ( Fig 1E ) , indicating that locomotor activity of forebrain-specific Rer1 cKO mice was normal . The brain weight and size of cortex from the forebrain-specific Rer1 cKO brain appeared 20–30% smaller than that of control mice ( Fig 1F and 1G ) . We stained sagittal sections of the brains from 5-month-old control and forebrain-specific Rer1 cKO mice with hematoxylin and eosin ( H&E ) and found that forebrain-specific Rer1 cKO mice had a smaller cerebral size than control mice ( Fig 1H ) . The forebrain-specific Rer1 cKO mice had a thinner cerebral cortex with a lower cell density and a smaller hippocampus than control mice ( Fig 1I ) . Since neurodegeneration is often accompanied by inflammatory responses such as gliosis , we analyzed the neuronal and astrocyte populations in control and forebrain-specific Rer1 cKO mice brains . We first prepared the lysates of the cerebral cortex from 5-week-old control and Rer1 cKO mice and examined the protein levels of a neuronal marker , NeuN , and an astrocyte marker , GFAP via western blot analysis ( Fig 1J ) . NeuN protein levels in forebrain-specific Rer1 cKO mice was slightly but significantly lower than that of control mice ( P < 0 . 05 ) ( Fig 1K ) . In contrast , GFAP protein levels apparently increased in forebrain-specific Rer1 cKO mice at 5 weeks postpartum , compared with that in control mice ( P < 0 . 01 ) ( Fig 1K ) . We also examined the cerebral cortex of 5-month-old control and forebrain-specific Rer1 cKO mice via immunohistochemical staining with anti-GFAP antibody ( Fig 1L ) . GFAP immunoreactivity increased by approximately 2 folds in the cerebral cortex of forebrain-specific Rer1 cKO mice ( P < 0 . 01 ) ( Fig 1M ) , suggesting that Rer1 loss causes prominent gliosis in the mouse brain . On postnatal day 0 ( P0 ) , brain size was comparable between forebrain-specific Rer1 cKO mice and controls ( Rer1+/inv-trap , Rer1inv-trap/inv-trap , Rer1+/inv-trap; Emx-Cre ) ( Fig 2A ) . However , the cellularity of the upper cortical layer ( especially layer II/III ) was reduced in the forebrain-specific Rer1 cKO brain at P0 ( Fig 2B ) . We also evaluated cortical malformations via immunostaining , using antibodies against layer markers , Cux1 ( layers II-IV ) , Ctip2 ( layer V ) , and Tbr2 ( intermediate progenitors in the subventricular zone , SVZ ) . The Cux1 signal decreased in the forebrain-specific Rer1 cKO brain ( Fig 2C ) . In contrast , Ctip2- and Tbr2-positive neurons were observed to a similar extent in control and forebrain-specific Rer1 cKO cerebral cortex ( Fig 2D and 2E ) . We also examined GFAP expression in the cerebral cortex of these mice on E18 . 5 via immunohistochemistry ( IHC ) ( Fig 2F ) and found no significant difference between control and the forebrain-specific Rer1 cKO brain on E18 . 5 ( Fig 2G ) , suggesting that switching from neurogenesis to gliogenesis normally takes place during the brain development . We further conducted western blot analysis of neuronal markers and markers for gliosis , ER stress , and apoptosis in cerebral lysates on E18 . 5 and quantitated their levels ( Fig 2H ) . Cux1 level was significantly reduced in the forebrain-specific Rer1 cKO brain ( P < 0 . 0001 ) ( Fig 2I ) , but Ctip2 ( Fig 2J ) was not , suggesting that the generation of layer ( II–IV ) neurons was impaired by Rer1 loss . The neuronal progenitor marker , Tbr2 ( Fig 2K ) and the astrocyte marker , GFAP ( Fig 2L ) levels were almost unchanged in the Rer1-deficient cerebral cortex . Protein levels of BiP , a representative ER chaperone , in the Rer1-deficient cerebral cortex was slightly increased compared to that in the control mice ( P < 0 . 01 ) ( Fig 2M ) , suggesting that the loss of Rer1 activates ER stress in the brain to some extent . In contrast , the amount of cleaved caspase-3 levels , which indicates apoptosis rate , were almost unchanged ( Fig 2N ) . These data suggest that the cortical malformation in the Rer1-deficient cerebral cortex occurs independently of neuronal cell death . Since Rer1 is involved in the γ-secretase complex formation , we investigated whether γ-secretase activity was affected in the cerebrum of forebrain-specific Rer1 cKO mice on E18 . 5 . We examined γ-secretase activity via an in vitro assay , using an intramolecularly quenched fluorogenic peptide probe , which contains a C-terminal β-APP amino acid sequence cleaved by γ-secretase . We incubated membrane extracts from the cerebrum of control or forebrain-specific Rer1 cKO mice with this probe overnight at 37°C and then analyzed γ-secretase activity . In Rer1-deficient cerebrum lysates , γ-secretase activity was reduced to 80% of that in the control lysate ( P < 0 . 05 ) ( Fig 3A ) . We also assessed γ-secretase activity and Wnt signaling in the cerebrum by monitoring the Notch intracellular domain ( NICD ) and the active form of β-catenin ( dephosphorylated on Ser37 or Thr41 ) , respectively ( Fig 3B ) . Notch is one of the known substrates of γ-secretase and plays an essential role in Notch signaling , which determines the cell fate during the development of various tissues [24 , 25] . First , γ-secretase triggers Notch signaling by cleaving the intramembrane domain of Notch . The cleaved Notch intracellular domain is then translocated to the nucleus to suppress the expression of proneural genes , which regulate neural differentiation , thereby inhibiting neuronal differentiation . The level of NICD in the Rer1-deficient cerebrum was significantly decreased to 80% of that in the control lysates ( P < 0 . 01 ) ( Fig 3C ) . In contrast , β-catenin protein levels were comparable between E18 . 5 control and Rer1-deficient cerebral cortex , indicating that Wnt signaling is superficially unaffected by Rer1 loss ( Fig 3B and 3C ) . These results suggest that Rer1 loss reduces γ-secretase activity and Notch signaling in the cerebrum . We further confirmed that Notch signaling is reduced during cerebral cortex development of Rer1-deficient mice by staining brain sections from control and forebrain-specific Rer1 cKO mice on E13 . 5 , using an anti-NICD antibody ( Fig 3D ) . A strong NICD signal was observed in the SVZ/intermediate zone ( IZ ) , while a weak signal was observed in the ventricular zone ( VZ ) of the control cerebrum; however , the NICD signal was apparently reduced in the Rer1 cKO cerebrum . We also examined the amount of NCT , NICD , and active-β-catenin in E13 . 5 mouse cerebral cortex lysates from control and Rer1 cKO mice to monitor the γ-secretase complex formation , Notch signaling , and Wnt signaling , respectively ( Fig 3E ) . NCT levels decreased in cerebral cortices of Rer1-deficient mice compared with that of control mice . Consistently , NICD levels decreased in the forebrain-specific Rer1 cKO cerebrum , suggesting that γ-secretase activity and Notch signaling are downregulated in the absence of Rer1 . In contrast , active-β-catenin protein levels were comparable in the E13 . 5 control and Rer1-deficient cerebral cortex , indicating that Wnt signaling is unaffected upon Rer1 loss . Furthermore , Hes1 , one of the downstream transcription factors in Notch-signaling pathway , was significantly down-regulated in the forebrain-specific Rer1 cKO mice ( P < 0 . 05 ) ( Fig 3F ) . Together , these results suggest that Rer1 loss decreases γ-secretase activity , thereby impairing Notch signaling , resulting in impaired cortical development in the mouse cerebrum . Since Notch signaling plays a pivotal role in neurogenesis by regulating neural stem cell ( NSC ) maintenance , we investigated the population of NSCs in the Rer1-deficient cerebral cortex on E13 . 5 , where neurogenesis occurs actively . We examined the proliferative potential of NSCs or progenitor cells via immunostaining , using an antibody against phosphorylated histone H3 ( pHH3 ) , which is highly expressed during cell division of neural stem or progenitor cells ( Fig 3G ) . The number of pHH3-positive dividing cells significantly decreased in the Rer1-deficient cerebral cortex , especially in the apical region ( P < 0 . 05 ) ( Fig 3H ) . We also performed EdU incorporation assay to determine the number of proliferating cells in control and forebrain-specific Rer1 cKO cerebral cortex on E13 . 5 ( Fig 3I ) . The number of proliferating cells decreased significantly in the forebrain-specific Rer1 cKO cerebral cortex relative to control mice ( P < 0 . 01 ) ( Fig 3J ) . Thereafter , we performed IHC , using anti-Pax6 antibody to assess the effect of Rer1 deficiency on NSCs ( Fig 3K ) . The number of Pax6-positive cells was significantly reduced in the apical area of cerebral cortex of the forebrain-specific Rer1 cKO cerebral cortex compared to control mice , although the total number of Pax6-positive cells in the total cerebral cortex area was comparable between them ( P < 0 . 001 ) ( Fig 3L ) , suggesting that Rer1 is necessary for the maintenance of normal pool of neural stem cells in the ventricular zone . We further investigated whether Rer1 is required for NSC maintenance . NSCs form colonies referred to as neurospheres , in vitro . We harvested NSCs from the brains of control and forebrain-specific Rer1 cKO mice on E14 . 5 to evaluate their colony ( neurosphere ) formation activity from a single NSC ( Fig 3M ) . NSCs harvested from E14 . 5 control mice efficiently formed neurospheres ( average number of 16 colonies from 200 cells ) , indicating their normal self-renewal . In contrast , NSCs from Rer1-deficient mice hardly formed neurospheres ( P < 0 . 01 ) ( Fig 3N ) , thereby supporting the importance of Rer1 in maintaining the pool of NSCs . To investigate the effect of Rer1 depletion on γ-secretase complex formation and activity , we examined Rer1-deficient HAP1 cells generated using a CRISPR/Cas-mediated genome editing system . Immunoblot analysis using anti-Rer1 antibodies confirmed the absence of Rer1 protein in Rer1-deficient HAP1 cells ( Fig 4A ) . Interestingly , NCT , PS1 , and PEN2 protein levels in Rer1-deficient HAP1 cells decreased significantly to 60–70% of those in wild-type cells ( NCT and PS1: P < 0 . 001 , PEN2: P < 0 . 01 ) ( Fig 4A and 4B ) . In addition , a hypoglycosylated form of NCT was observed in Rer1-deficient HAP1 cells , as reported previously [19] . We also investigated the effects of Rer1 depletion on other plasma membrane proteins such as LRP6 , integrin , pan-cadherin , and Slc3A2 . The levels of these proteins were unaffected in Rer1-deficient cells ( Fig 4C ) , suggesting that the overall biosynthesis of plasma membrane proteins is not severely impaired in Rer1-deficient cells . Furthermore , we investigated whether Rer1 re-expression could recover protein levels of γ-secretase subunits in Rer1-deficient cells . We introduced GFP or GFP-Rer1 into Rer1-deficient cells , using a retroviral vector system and examined the effect of Rer1 re-expression on γ-secretase subunit levels via immunoblot analysis ( Fig 4D ) . The expression of GFP-Rer1 in Rer1-deficient cells increased NCT and PS1 levels by 1 . 2 folds compared with that in control cells expressing GFP alone ( P < 0 . 001 ) ( Fig 4E ) . We also examined the cell surface levels of NCT and PS1 in wild-type and Rer1-deficient cells via a cell surface biotinylation assay ( Fig 4F ) . We labeled cell surface proteins with biotin and pulled them down using avidin beads to detect γ-secretase levels at the plasma membrane . Immunoblot analysis indicated that the amount of NCT and PS1 on the cell surface of Rer1-deficient cells was reduced to approximately 60% of that in wild-type cells ( NCT: P < 0 . 05 , PS1: P < 0 . 01 ) ( Fig 4G ) . However , the ratio of cell surface NCT and PS1 to total proteins was comparable in wild-type and Rer1-deficient cells ( Fig 4H ) . To investigate whether NCT assembles with PS1 in Rer1-deficient cells , we immunoprecipitated γ-secretase complexes from wild-type and Rer1-deficient cell lysates , using an anti-NCT antibody and examined the immunoprecipitates upon immunoblotting with antibodies against PS1 and NCT ( Fig 4I ) . The association of NCT and PS1 was detected in Rer1-deficient cells , suggesting that γ-secretase subunits can form a complex even in the absence of Rer1 . We also examined the gene expression of γ-secretase components via quantitative PCR analysis . The expression of PS1 and NCT in Rer1-KO HAP1 cells was comparable to that in control cells , suggesting that gene expression of γ-secretase components was not largely affected by the loss of Rer1 ( Fig 4J ) . To investigate the role of Rer1 in the stability of the γ-secretase complex , we examined the turnover rates of NCT and PS1 in wild-type and Rer1-deficient cells via treatment with a protein synthesis inhibitor , cycloheximide ( CHX ) . The γ-secretase complex is known to be formed by sequential assembly of APH-1 , NCT , PS1 and PEN2 in order [26] . PS1 is then processed endoproteolytically to a 28-kDa N-terminal fragment and a 17-kDa C-terminal fragment [27] . We used an anti-PS1 antibody , which can detect a C-terminal fragment of PS1 ( PS1 CTF ) . Full-length PS1 is hardly detected at a steady state because it is quickly degraded via a proteasome-dependent mechanism or endoproteolysis [26 , 28] , whereas PS1 CTF stably exists in a complex . NCT and PS1 CTF largely remained in wild-type cells at 24 h after CHX treatment ( Fig 5A ) . By contrast , levels of both proteins started to decrease at 5 h after CHX treatment in Rer1-deficient cells . Approximately 50% of NCT and PS1 were degraded within 24 h in Rer1-deficient cells compared with wild-type cells , in which most of both proteins remained at 24 h ( Fig 5B ) . Furthermore , we investigated whether γ-secretase subunits were degraded by the proteasome or lysosomes in Rer1-deficient cells . We treated wild-type and Rer1-deficient cells with a potent inhibitor of ER-associated protein degradation ( ERAD ) , eeyarestatin I ( EerI ) , or a proteasome inhibitor , MG132 , and investigated the involvement of the proteasome in the degradation of γ-secretase subunits ( Fig 5C and S4 Fig ) . ERAD inhibition significantly increased PS1 CTF levels in wild-type and Rer1-deficient cells ( P < 0 . 05 and P < 0 . 0001 , respectively ) , as reported previously [28] ( Fig 5D ) . NCT levels also tended to increase slightly but not significantly in the presence of EerI ( Fig 5E ) , suggesting that PS1 CTF and NCT are partially degraded by the proteasome in wild-type and Rer1-deficient cells . Since γ-secretase subunits began to decrease 5 h after CHX treatment in Rer1-deficient cells , we assumed that NCT and PS1 CTF might be partly transported to and degraded in lysosomes . To test this hypothesis , we treated Rer1-deficient cells with bafilomycin A1 ( BafA1 ) , which blocks lysosomal degradation by inhibiting V-ATPase . NCT and PS1 CTF protein levels significantly increased in BafA1-treated Rer1-deficient cells compared with that in vehicle ( DMSO ) -treated cells ( Fig 5F and 5G ) . The effects of BafA1 on NCT and PS1 CTF accumulation were alleviated in GFP-Rer1-expressing Rer1-deficient cells ( Fig 5H and 5I ) . Furthermore , we investigated whether some γ-secretase subunits are transported to lysosomes in Rer1-deficient cells . We incubated wild-type and Rer1-deficient cells with or without BafA1 and observed the subcellular localization of NCT via immunostaining with antibodies against NCT and a lysosomal marker Lamp1 ( Fig 5J ) . Endogenous NCT mainly localized to the ER in a meshwork pattern and partly to punctate structures in wild-type cells treated with or without BafA1 . Lamp1-positive lysosomes tended to aggregate in a cellular region in HAP1 cell lines in the presence of BafA1 . NCT-positive puncta subtly increased in Rer1-deficient cells at a steady state and prominently increased in Rer1-deficient cells treated with BafA1 . Approximately 50% of NCT-positive puncta localized close to the Lamp1-positive lysosomes or overlapped with them in BafA1-treated Rer1-deficient cells ( Fig 5K ) , indicating that NCT is partly transported to lysosomes . These results suggest that immature γ-secretase subunits or complexes are no longer retained in the ER and are mislocalized to lysosomes for degradation in the absence of Rer1 . We also examined γ-secretase complex formation via Blue Native PAGE ( BN-PAGE ) ( Fig 5L ) . The amount of γ-secretase complex was drastically reduced in Rer1-KO cells compared to that in control cells; however , complex formation still occurred to some extent under this condition . Notably , the reduction in γ-secretase complex was recovered upon addition of BafA1 , which inhibits lysosomal function , suggesting that γ-secretase subunits are largely targeted to and degraded in lysosomes in the absence of Rer1 . Furthermore , NCT was underglycosylated in Rer1-deficient cells , as reported previously [19] , in Rer1-knockdown cells ( Fig 5M ) , suggesting that Rer1 is required for the cell surface expression of fully matured γ-secretase complex .
The present study shows that Rer1 exerts positive effects in the maintenance of γ-secretase activity by mediating proper expression of γ-secretase complex on the cell surface . Additionally , Rer1 is involved in the proliferation of NSCs during cerebral cortex development by modulating the onset of Notch signaling , which regulates cortical development and higher brain function . These findings suggest that Rer1 functions as the early-Golgi quality control , which mediates normal expression and functioning of the γ-secretase complex during mammalian brain development . Rer1 has been proposed as a limiting factor , which negatively regulates the assembly of γ-secretase subunits , since siRNA-mediated knockdown of Rer1 facilitates γ-secretase assembly and cell surface expression in cultured cells [19] . However , we observed that γ-secretase activity in the brain lysates was significantly decreased in Rer1-deficient mice . In addition , we found that a complete loss of Rer1 significantly reduced γ-secretase components on the cell surface probably because of their mistargeting to lysosomes for degradation . These results raise a question regarding the different effects of Rer1 knockdown and Rer1 deletion on expression of the γ-secretase complex . In both cases , some of the γ-secretase subunits or complexes are transported to post-Golgi compartments similarly both in Rer1-knockdown and knockout cells; however , their destinations are different . One possible explanation is that late-Golgi or post-Golgi quality control systems function more strictly in HAP1 cells and mice than in other cultured cell lines used in previous studies and deliver immature γ-secretase subunits or complexes to lysosomes . Alternatively , a small amount of Rer1 in Rer1-knockdown cells might be sufficient to support the maturation of γ-secretase subunits or complexes , allowing delivery to the cell surface . By contrast , a prolonged and profound lack of Rer1 may allow for leakage of γ-secretase subunits or complexes from the ER , resulting in their lysosomal targeting . Our observations suggest that Rer1 exerts positive effects on the cell surface expression of γ-secretase . In yeast , Rer1p is required for accurate formation of an iron transporter complex consisting of Fet3p and Ftr1p by retaining unassembled Fet3p in the ER [7] . In the absence of Rer1p , unassembled Fet3p is mislocalized to vacuoles and degraded there . Ftr1p also becomes unstable and degraded probably by ER-associated degradation in Rer1-deficient cells , suggesting a positive role of Rer1p in the formation of the iron transporter complex . Furthermore , Rer1 loss leads to lysosomal degradation of unassembled components such as the γ-subunit of acetylcholine receptor and immature rhodopsin in mammalian cells [6 , 22] . These observations support our hypothesis that Rer1 mediates the assembly or maturation of membrane protein complexes as the early-Golgi quality control system by retaining unassembled components until they form an appropriately matured complex . Interestingly , some of the γ-secretase components and/or complex were targeted to lysosomes for degradation in Rer1-deficient cells , suggesting the existence of late-Golgi or post-Golgi quality control system [4 , 29 , 30] . Multiple quality control systems through the exocytosis pathway would allow for efficient formation of membrane protein complexes and their cell surface expression at a proper timing . NSCs harvested from Rer1-deficient cerebral cortex hardly form neurospheres , indicating that self-renewal of NSCs is partly impaired by Rer1 loss . In addition , Rer1 deficiency in the developing mouse cerebrum decreased the number of mature neurons , including Cux1-positive upper layer neurons in the cortex , and reduced the size of the cerebral cortex ( Figs 1–3 ) . In such mice , the number of dividing NSCs decreased in the cerebral cortex . These results suggest that Rer1 loss impairs the proliferation of NSCs , which finally leads to the exhaustion of neuronal progenitor cells . Interestingly , these phenotypes of Rer1-deficient mice are reminiscent of those of mice defective in γ-secretase and Notch signaling [31–33] . γ-Secretase triggers Notch signaling , which is known to regulate the expansion of the neuronal progenitor pool for appropriate regulation of brain size . In this study , we demonstrated that Notch signaling was significantly reduced in the Rer1-deficient cerebrum probably because of a reduced activity of γ-secretase complex . These findings suggest that Rer1 maintains Notch signaling by supporting proper γ-secretase complex expression to supply a sufficient progenitor pool for cortical development . Recently , it has been reported that Rer1 is also involved in the assembly and transport of voltage-gated sodium channels in Purkinje cell [34] . Thus , Rer1 could also affect other membrane protein complexes such as voltage-gated sodium channels or the activity of other γ-secretase substrates in addition to Notch1 to regulate brain development and behaviors . In humans , Rer1 , located in 1p36 region , is significantly associated with 1p36-deletion syndrome . 1p36-deletion syndrome is caused by a large deletion in the subtelomeric region of chromosome 1 , and it occurs in 1 in 5000 people . Patients with this syndrome exhibit a brain disorder with mental retardation and behavioral disorder . In addition , they also have structural brain abnormalities , including a large anterior fontanelle and microcephaly . Notably , Rer1 loss in the cerebral cortex resulted in cerebral hypoplasia and behavioral disorder ( Fig 1 ) , which recapitulates the brain defects present in most patients with 1p36-deletion syndrome , suggesting that Rer1 mutations are one of the causative factors for this syndrome . However , the symptoms in patients result from haploinsufficiency due to 1p36 microdeletion . Since Rer1 heterozygous mice did not develop the striking phenotypic abnormalities observed in forebrain-specific Rer1-deficient mice , heterozygous deletion of the other genes would also contribute to brain anomalies observed in 1p36-deletion syndrome . Interestingly , the downstream targets of Notch signaling such as Hes4 and Hes5 are also present in this region [35] , although over 90 genes are contained in the 1p36 microdeletion . Reduced expression of such genes together with Rer1 in patients may cause abnormal brain development . Further investigation of γ-secretase activity and Notch signaling in patients with 1p36-deletion syndrome will be important for establishing the involvement of Rer1 and the pathogenic mechanisms underlying the disease etiology .
All animal procedures were performed in accordance with the guidelines of the Animal Care and Experimentation Committee of Gunma University , and all animals were bred at the Institute of Animal Experimental Research of Gunma University . FlipRosaβgeo-trapped embryonic stem ( ES ) cell clones no . EUCJ0172c09 ( EUCOMM ) were used to generate heterozygous mice ( Rer1+/trap ) , using standard protocols ( S1 and S2 Figs ) . The FlipRosaβgeo cassette comprises a conventional gene trap element and pairs of inversely oriented heterotypic recombinase target sites ( RTs ) , such as loxP and FRT sites that flank the gene trap element ( S3 Fig ) . FLPe and Cre recombinases can invert the gene trap element of FlipRosaβgeo flanked by RTs via directional site-specific recombination , thereby first repairing and then re-inducing the gene trap mutation [36] . To generate the mouse ( Rer1+/inv-trap ) harboring FLPe-inverted gene trap insertions , in which the gene trap mutation for Rer1 expression is invalidated , we crossed Rer1+/trap mice with actin-flippase transgenic mice ( B6; SJL-TG ( ACTFLPe ) 9250Dym/J ) ( Jackson Laboratory ) . To generate forebrain-specific Rer1-deficient mice , Emx1-Cre transgenic mice [23] were crossed with the Rer1+/inv-trap mice to re-induce the gene trap mutation via Cre recombinase-mediated inversion in the forebrain . Rer1-deficient HAP1 cell lines , edited by CRISPR/Cas to contain a frame shift mutation in a coding exon 3 of Rer1 , were obtained from horizon discovery ( Horizon Discovery ) . The guide RNA sequence was as follows: 5’-CACCCTACACGGCTGTGCGA-3’ . HAP1 cells were cultured in Iscove’s Modified Dulbecco’s Medium ( IMDM ) supplemented with 10% fetal calf serum and penicillin/streptomycin ( Wako ) in a 5% CO2 incubator at 37°C . HAP1 cells expressing GFP or GFP-Rer1 were generated via retroviral transduction . To generate retroviruses , Plat-E cells were co-transfected with pMXs-IP-GFP-mouse Rer1 and pCG-VSV-G using FuGENE HD ( Promega ) . HAP1 cells were then infected with the recombinant retroviruses and selected in medium containing 3 μg/mL puromycin , as reported previously [37] . MG132 , γ-secretase inhibitor ( L685 , 458 ) , and a fluorescence-quenching substrate for γ-secretase ( Nma-Gly-Gly-Val-Val-Ile-Ala-Thr-Val-Lys ( Dnp ) -D-Arg-D-Arg-D-Arg-NH2 ) were purchased from the Peptide Institute , Japan . Protease inhibitor cocktail ( complete EDTA-free protease inhibitor ) and eeyarestatin I were purchased from Roche , USA . Bafilomycin A1 was from Wako Pure Chemical , Japan; cycloheximide from MP Biomedicals , USA; CHAPSO from Dojin , Japan; and Sulfo-NHS-Biotin from Thermo Fisher Scientific , USA . E18 . 5 mouse cerebral cortices were homogenized using Dounce homogenizer in homogenate buffer ( 20 mM HEPES , pH 7 . 5 , 50 mM KCl , 2 mM EGTA ) with a protease inhibitor cocktail . Lysates were centrifuged at 4°C , 800 × g for 10 min . The resulting supernatants were centrifuged at 4°C , 100 , 000 × g for 1 h . The membrane pellet was resuspended in a buffer containing 20 mM HEPES pH 7 . 0 , 150 mM KCl , 2 mM EGTA , 1% CHAPSO , and protease inhibitor cocktail . γ-Secretase activity was measured by incubating solubilized membranes with the fluorescence-quenching substrate for γ-secretase overnight at 37°C in the absence or presence of L-685 , 458 , as reported previously [38] . Mouse brains were homogenized in ice-cold homogenate buffer ( 50 mM Tris-HCl pH 7 . 4 , 0 . 25 M sucrose , 1 mM EDTA ) with a protease inhibitor cocktail , using the mini homogenizer tube BioMasher II ( Nippi ) . After homogenization , an equal amount of lysis buffer ( 50 mM Tris-HCl pH 7 . 4 , 250 mM NaCl , 1% Triton X-100 , 0 . 1% SDS , 0 . 2% sodium deoxycholate , and 1 mM EDTA ) was added and the lysates were incubated on ice for 30 min . The lysates were centrifuged at 4°C , 15 , 000 rpm for 15 min . Supernatants were subjected to immunoblot analysis using specific antibodies . HAP1 cells were lysed in a cell lysis buffer ( 50 mM Tris-HCl pH 7 . 4 , 0 . 2% SDS , 1% sodium deoxycholate ) with a protease inhibitor cocktail and 1 mM phenylmethanesulfonyl fluoride ( PMSF ) . The lysates were centrifuged at 4°C , 15 , 000 rpm for 15 min and the supernatants were then subjected to immunoblot analysis using specific antibodies . HAP1 cells were cultured on coverslips and fixed with 3% paraformaldehyde in PBS for 10 min . The cells were permeabilized with 0 . 1% Triton X-100 and incubated with PBS containing 5% NDS for 1 h for blocking , and then treated with specific antibodies . Images were acquired using an FV1000 confocal microscope ( Olympus ) with a 100× PlanApo oil immersion lens ( 1 . 40 numerical aperture; Olympus ) . The following primary antibodies were used for immunoblotting: anti-RER1 ( Sigma , R4407 ) , anti-Nicastrin ( Sigma , N1660 ) , anti-Presenilin-1 [D39D1] ( Cell Signaling Technologies , 5643 ) , anti-Presenilin-1 [PS1loop] ( Millipore , MAB5232 ) , anti-PEN2 ( Cell Signaling Technologies , 5451 ) , anti-CD49b ( BD , 611016 ) , anti-LRP6 ( Cell Signaling Technologies , 2560 ) , anti-Pan-cadherin ( Cell Signaling Technologies , 4068 ) , anti-CD98 ( Santa Cruz , sc-9160 ) , anti-GFP ( Fitzgerald , RDI-GRNFP3abg ) , anti-actin [C4] ( Millipore , MAB1501 ) , anti-NeuN [A60] ( Millipore , MAB377 ) , anti-GFAP ( Frontier Institute , GFAP-Rb-Af800 ) , anti-Cux1 ( Santa Cruz , sc13024 ) , anti-Ctip2 ( Abcam , ab18465 ) , anti-Tbr2 ( Abcam , ab23345 ) , anti-NICD [D3B8] ( Cell Signaling Technologies , 4147 ) , anti-Notch1 [D1E11] ( Cell Signaling Technologies , 3608 ) , anti-β-catenin ( BD , 610153 ) , anti-active-β-catenin [8E7] ( MERCK , 05–665 ) , and anti-cleaved caspase-3 ( Cell Signaling Technology , 9661 ) antibodies . The following primary antibodies were used for immunostaining: anti-Nicastrin ( Sigma , N1660 ) , anti-Lamp1 [H4A3] ( Santa Cruz , sc20011 ) , anti-phospho-Histone H3 [Ser10] ( Cell Signaling Technologies , 9701 ) , anti-GFAP ( Frontier Institute , GFAP-Rb-Af800 ) , anti-Cux1 ( Santa Cruz , sc13024 ) , anti-Ctip2 ( Abcam , ab18465 ) , anti-Tbr2 ( Abcam , ab23345 ) , anti-Pax6 ( BioLegend , PRB-278P ) and anti-NICD [D3B8] ( Cell Signaling Technologies , 4147 ) antibodies . The following secondary antibodies were used for immunostaining: goat anti-rabbit Alexa Fluor-568 , goat anti-mouse Alexa Fluor-647 , donkey anti-rabbit Alexa Fluor-594 , and donkey anti-rat Alexa Fluor-594 ( all from Life Technologies ) antibodies . Rabbit anti-Rer1 antibodies were generated against the C-terminal regions of mouse Rer1 ( NH2-C+KRRYKGKEDVGKTFAS-COOH coupled to KLH; TK craft corp . ) . Pregnant female mice were intraperitioneally administered 50 mg kg-1 body weight of EdU . After 2 h , E13 . 5 embryos were fixed in 4% paraformaldehyde ( PFA ) in phosphate-buffered saline ( PBS ) . Subsequently , brain tissues were immersed in 15% and then 30% sucrose in PBS and frozen for sectioning . EdU detection was performed in accordance with the manufactures’ Click-it EdU Alexa 594 imaging kit protocol ( Thermo Fisher Scientific ) . Adult mice were transcardially perfused with 4% paraformaldehyde in phosphate buffer ( pH 7 . 4 ) , as described previously [39] . Brain tissues from embryos were fixed in the same fixative solution overnight . Thereafter , tissues were embedded in paraffin , sectioned , and then stained using Meyer's H&E . For immunohistochemistry for paraffin-embedded tissue , tissue sections were subjected to antigen retrieval with a microwave oven in 0 . 01 M citrate buffer ( pH 6 . 0 ) for 10 min . After blocking with 3% H2O2 in MeOH for 30 min and then 5% BSA in PBS for 30 min , sections were incubated with primary antibodies , followed by incubation with EnVision Plus System-HRP Labelled Polymer Anti-Rabbit kit ( K4003 , DakoCytomation ) . The signal was visualized using Liquid DAB substrate chromogen system ( DakoCytomation ) . Hematoxylin was used as a counterstain . For immunohistochemical analysis of cryosections , fixed tissues were cryoprotected via sequential immersion in 15% and 30% sucrose in PBS overnight and were embedded in Tissue-Tek OCT compound ( Sakura Finetek ) . After embedding , 10-μm cryosections were cut and immunolabeled . The tissue sections were then subjected to antigen retrieval in Histo VT One ( NACALAI TESQUE ) at 70°C for 10 min ( for pHH3 and Pax6 staining ) , target retrieval solution S1700 ( DakoCytomation ) at 105°C for 15 min ( for NICD staining ) , target retrieval solution S1700 ( DakoCytomation ) at 90°C for 5 min ( for Cux1 staining ) , and 0 . 01 M citrate buffer ( pH 6 . 0 ) at 90°C for 10 min ( for Tbr2 , Ctip2 , and GFAP staining ) . After blocking with 5% BSA in PBS for 30 min , sections were incubated with primary antibodies , followed by incubation with fluorescently labeled secondary antibodies . The stained sections were analyzed via confocal microscopy ( FV1000 , Olympus ) or fluorescence microscopy ( BZ-9000 , Keyence ) . GFAP- , Pax6- , or EdU-positive cells were enumerated using BZ-X analyzer software ( Keyence ) . Behavioral tests were performed with male 3-month-old littermates . The open-field test and rotarod test were performed , as described previously [40] . All tests used equipment from O’Hara & Co . Ltd . Neural cells were isolated from E14 . 5 cerebral cortex . The isolated cells were plated at a density of 2 × 105 cells/mL in KBM Neural Stem Cell ( KOHJIN BIO #16050100 ) with KBM neural stem cell supplement containing epidermal growth factor and fibroblast growth factor ( KOHJIN BIO ) and cultured at 37°C in 5% CO2 . After 1 week , neurospheres were collected and dissociated by TrypLE ( ThermoFisher Scientific ) , and 200 dissociated cells were plated in each well of a 96-well ultra-low attachment plate ( Corning ) for secondary sphere formation . The secondary neurospheres were enumerated after 1 week of culture . Cells were washed with ice-cold HBSS twice ( Gibco ) and incubated for 30 min on ice with 0 . 5 mg/mL Sulfo-NHS-Biotin ( Thermo Fisher Scientific ) diluted in PBS . After three washes with HBSS , cells were incubated with 50 mM NH4Cl for 5 min to quench excess biotin . Cells were then lysed with a lysis buffer ( 50 mM Tris-HCl pH 7 . 5 , 150 mM NaCl , 0 . 5 mM EDTA , 1% Triton X-100 , Protease Inhibitor Cocktail ) . Cell lysates were pulled down with streptavidin agarose resin ( Thermo Fisher Scientific ) overnight . Pulled down samples were washed twice with lysis buffer and then subjected to immunoblot analysis , using specific antibodies . HAP1 cells were prepared with a native sample buffer ( Thermo Fisher Scientific ) , containing 0 . 5% DDM ( n-dodecyl β-ᴅ-maltoside ) and a protease inhibitor mixture and subjected to Blue Native PAGE , using the Novex Bis-Tris gel system ( Thermo Fisher Scientific ) in accordance with the manufacturer’s instructions . Total RNA was extracted using the RNeasy mini kit ( Qiagen ) and subjected to RT with a ReverTra Ace qPCR RT kit ( Toyobo ) . The resulting cDNA was then subjected to real-time PCR analysis with 1 x SYBR Green PCR master mix ( Takara ) and gene specific primers . Assays were performed with Thermal Cycler Dice Real Time System III ( Takara ) . The sequence of the various primers ( the forward primer and the reverse primer , respectively ) were 5′-GGACCCGAGAAGACCTCCTT-3′ and 5′-GCACATCACTCAGAATTTCAATGG-3′ for mouse acidic ribosomal phosphoprotein , 5′-TATCATGGAGAAGAGGCGAAGG-3′ and 5′-TTCTCTAGCTTGGAATGCCGG-3′ for mouse Hes1 , 5′-GAAGGTGAAGGTCGGAGTCA-3′ and 5′-TGGACTCCACGACGTACTCA-3′ for human GAPDH , 5′-AGCAGTATCCTCGCTGGTGA-3′ and 5′-TGAAATCTCCCAATCCAAGTTT-3′ for human PSEN1 , 5′-CAGTGGCTTCCTTTGTCACC-3′ and 5′-GAGCTGCCAATGTAGTCAAAAG-3′ for human NCSTN . HAP1 cells were treated with 10 μg/mL CHX , and the cell extracts were prepared at specified time points and subjected to immunoblot analysis with the indicated antibodies . HAP1 cells were treated with 5 μM MG132 for 2 h , 2 . 5 μM eeyarestatin I , or 200 nM BafA1 for 16 h . Cell lysates were then subjected to immunoblot analysis with indicated antibodies . Data were analyzed using GraphPad Prism 6 ( GraphPad Software , Inc . , USA ) , and expressed as mean ± standard error of the mean values . Two-tailed Student’s t-tests and two-way analysis of variance was used to evaluate significance and calculate P values . P values less than 0 . 05 were considered statistically significant .
|
We showed that Rer1 functions as an early-Golgi quality control pathway that maintains γ-secretase activity by maintaining sufficient cell surface expression of γ-secretase complex during cerebral cortex development , thereby modulating Notch signaling .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"lysosomes",
"molecular",
"probe",
"techniques",
"brain",
"notch",
"signaling",
"membrane",
"proteins",
"molecular",
"biology",
"techniques",
"immunologic",
"techniques",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"analysis",
"methods",
"specimen",
"preparation",
"and",
"treatment",
"staining",
"cerebral",
"cortex",
"immunoblot",
"analysis",
"cerebrum",
"molecular",
"biology",
"cell",
"membranes",
"signal",
"transduction",
"immunohistochemistry",
"techniques",
"immunostaining",
"anatomy",
"cell",
"biology",
"biology",
"and",
"life",
"sciences",
"cell",
"signaling",
"histochemistry",
"and",
"cytochemistry",
"techniques"
] |
2018
|
Rer1-mediated quality control system is required for neural stem cell maintenance during cerebral cortex development
|
Aplasia cutis congenita ( ACC ) manifests with localized skin defects at birth of unknown cause , mostly affecting the scalp vertex . Here , genome-wide linkage analysis and exome sequencing was used to identify the causative mutation in autosomal dominant ACC . A heterozygous Arg-to-His missense mutation ( p . R930H ) in the ribosomal GTPase BMS1 is identified in ACC that is associated with a delay in 18S rRNA maturation , consistent with a role of BMS1 in processing of pre-rRNAs of the small ribosomal subunit . Mutations that affect ribosomal function can result in a cell cycle defect and ACC skin fibroblasts with the BMS1 p . R930H mutation show a reduced cell proliferation rate due to a p21-mediated G1/S phase transition delay . Unbiased comparative global transcript and proteomic analyses of ACC fibroblasts with this mutation confirm a central role of increased p21 levels for the ACC phenotype , which are associated with downregulation of heterogenous nuclear ribonucleoproteins ( hnRNPs ) and serine/arginine-rich splicing factors ( SRSFs ) . Functional enrichment analysis of the proteomic data confirmed a defect in RNA post-transcriptional modification as the top-ranked cellular process altered in ACC fibroblasts . The data provide a novel link between BMS1 , the cell cycle , and skin morphogenesis .
Aplasia cutis congenita ( ACC [MIM 107600] ) manifests at birth as a localized skin defect that usually heals with a hypertrophic scar . Most commonly the scalp skin is affected and results in localized alopecia at the site of the defect , but sometimes the defect can extend into deeper structures involving the dura mater or osseous structures . ACC has to be distinguished clinically from birth trauma or intrauterine herpetic infections . Most reported cases are sporadic , but autosomal dominant inheritance has been reported as well [1] . A causative gene mutation for non-syndromic ACC has not been reported so far . While the majority of patients with aplasia cutis have no other congenital abnormalities , aplasia cutis can occur as part of rare syndromes with a wide spectrum of anomalies . For example , some patients with Johanson-Blizzard syndrome [MIM 243800] have been reported to have aplasia cutis of the scalp skin . This syndrome is characterized by nasal alar hypoplasia , hypothyroidism , pancreatic achylia and congenital deafness and is caused by a mutation in the UBR1 gene [2] . Adams-Oliver syndrome ( AOS [MIM 100300] ) is characterized by ACC and transverse limb defects , but a wide range of additional congenital anomalies have been reported in patients with AOS , including congenital heart defects . Recently gain-of-function mutations in the ARHGAP31 gene encoding a Cdc42/Rac1 regulatory protein have been reported in AOS [3] , and two more cases of AOS were reported to harbor homozygous mutations in the DOCK6 gene , encoding a guanidine nucleotide exchange factor that activates Cdc42 and rac1 [4] . A recent report showed mutations in the transcriptional regulator of the Notch pathway , RBPJ , in two small families with AOS [5] . Furthermore , mutations in the KCTD1 gene were recently reported in patients with Scalp-Ear-Nipple syndrome , which manifests with scalp ACC lesions as well [6] . However , the majority of individuals with aplasia cutis have no other congenital anomalies , suggesting a distinct pathomechanism between non-syndromic ACC and syndromic cases as in AOS or other syndromes in which aplasia cutis has been described . Histologically , ACC shows in most cases a complete absence of epidermis at birth suggesting a skin morphogenesis defect . It has been speculated that a defect in cell proliferation may be underlying ACC that may lead to a delay in skin closure ability late during development at an anatomic site where a steady expansion of the brain structures may require a rapid proliferation of the overlying skin . However , neither the genetic basis for ACC nor the pathomechanisms that result in the skin formation defects are known . Identifying the genetic causes for ACC and determining their molecular consequences promises to reveal new mechanisms that govern skin morphogenesis during development . Here , by combining genome-wide linkage analysis with exome sequencing approaches , a mutation in the ribosomal GTPase BMS1 is identified in autosomal dominant ACC that is associated with a p21-mediated G1/S phase cell cycle transition delay and results in a reduced cell proliferation rate . In addition , unbiased expression profiling and proteomic analyses in fibroblasts carrying this mutation independently confirm a central role of a p21-mediated G1/S phase cell cycle delay for the ACC phenotype .
In this study , a five-generation family with autosomal dominant inheritance of ACC was identified ( Fig . 1a ) . In this family the exclusive congenital anomaly is a localized absence of skin at the vertex and occipital area of the scalp that most often healed with a hypertrophic scar ( Fig . 1a ) . No other congenital anomalies were identified , excluding rare syndromes that can manifest with ACC [2] , [3] , [4] . Genome-wide linkage analysis revealed a single genomic region with a LOD score>2 on chromosome 10q11 . The maximal LOD score of 2 . 709 was reached with a sharp peak between marker rs1359280 located at 34 , 877 , 513 bp and marker rs7071514 located at 49 , 518 , 113 bp ( Fig . 1b ) ( an additional affected family member who carried the disease allele was born after the linkage study was completed , which further increased the LOD score ) . Haplotypes and recombination sites were independently confirmed using microsatellite markers . Whole-exome sequencing was performed in one affected individual of this family . Sequencing of coding regions was performed to a mean coverage of 151× to generate 17 . 3 Gb of sequence . Variants were filtered to exclude homozygous base pair changes , non-coding variants , synonymous variants and all non-synonymous changes that are present in dbSNP129 , the 1000genomes database and the NHLBI Exome Sequencing Project ( ESP ) database . This genome-wide sequencing approach resulted in the identification of a single non-synonymous heterozygous G>A base change within the linked genomic region . This sequence change was not identified in 5 , 351 unrelated individuals for which high-quality sequence calls were made at this position in the ESP sequence data , suggesting that the G>A sequence change is not a rare variant but pathogenic in this family with ACC . Bidirectional Sanger sequencing was performed in all members of this family , showing that none of the unaffected family members had the G>A sequence change , but all affected family members had this sequence change , thus co-segregating with the disease allele . Finally , sequencing of DNA from 100 geographically and ethnically matched unaffected control individuals did not reveal this mutation as well . This heterozygous G>A mutation results in an Arg-to-His amino acid change at position 930 ( p . R930H ) of a conserved Arginine within the C-terminal domain of the ribosome assembly GTPase BMS1 , which has previously not been implicated in skin morphogenesis ( Figure 1c–1e ) . BMS1 is a component of the U3 snoRNA-containing complex that has been shown to be essential in yeast and to be conserved in eukaryotes . Depletion of BMS1 interferes with pre-ribosomal RNA ( rRNA ) processing at sites A0 , A1 and A2 and the formation of the 40S ribosomal subunit [7] , [8] . Biochemical analyses demonstrated that BMS1 may function as a GTPase , with its GTPase activity located at the N-terminus of the protein . In addition to binding to U3 snoRNA , BMS1 binds the endonuclease Rcl1 in a GTP-dependent manner and BMS1 binding is required to recruit Rcl1 to preribosomes . Thus , it has been suggested that BMS1 functions as a GTP-regulated switch to deliver Rcl1 to preribosomes and has an essential role in the formation of the small ribosomal subunit [9] , [10] . The C-terminal domain of BMS1 has been proposed to act as a putative intramolecular GTPase-activating protein ( GAP ) domain and is separated from the rest of the protein by a flexible linker region [9] . However , the in vivo consequences of an impairment of this putative GAP domain for cellular functions have not been defined , and the BMS1 p . R930H mutation in ACC provides the first indication for a role of this domain for overall BMS1 function in vivo . BMS1 is ubiquitously expressed , including the proliferative developing skin of the scalp that is affected in ACC ( Fig . 1f , 1g ) . The BMS1 p . R930H mutation within the putative GAP domain does not affect BMS1 expression or its nucleolar localization ( Fig . S1 ) . Due to the heterozygous nature of the mutation within a regulatory domain of BMS1 with a remaining wild-type allele and the only localized congenital defect in ACC , it would be expected that the p . R930H mutation in BMS1 leads to rather modest abnormalities in pre-rRNA processing and therefore results in a restricted phenotype . Indeed , pulse-chase labeling experiments with ACCBMS1 ( p . R930H ) fibroblasts ( isolated from an affected member of this family ) and with matched unrelated wild-type fibroblasts , in which labeling of pre-rRNAs was performed with L-[methyl-3H]methionine , showed formation of all pre-rRNAs but with higher levels of persistent 45S pre-rRNA and the pre-rRNA band likely corresponding to the 30S pre-rRNA in ACCBMS1 ( p . R930H ) fibroblasts ( Fig . 2a–2c ) . Northern blotting experiments using RNA from ACCBMS1 ( p . R930H ) fibroblasts and control fibroblasts with probes for ITS-1 , ETS-1 and ITS-2 ( transcribed spacers ) showed an increase of 45S and 30S pre-rRNAs in mutant cells and a relative decrease of 21S and 18S-E pre-rRNAs ( precursors for 18S rRNA of the small ribosomal subunit ) , while 32S and 12S pre-rRNAs ( precursors for 5 . 8S and 28S rRNAs of the large ribosomal subunit ) were not affected ( Fig . 2d , Fig . S2 ) . The findings of the pulse-chase labeling and Northern blotting experiments in ACCBMS1 ( p . R930H ) fibroblasts are consistent with a delay in pre-rRNA processing affecting the small ribosomal subunit , as would be predicted based on the reported role of BMS1 for the small ribosomal subunit processome in yeast [7] , [8] . Similarly , inducible shRNA-mediated stable knockdown of BMS1 to achieve ∼40–50% of BMS1 transcript levels ( to correspond to heterozygous BMS1 cells ) resulted also in a relative increase of 45S and 30S pre-rRNAs , and a relative decrease in 21S and 18S-E pre-rRNAs , while 32S pre-rRNAs were reduced less when compared to the 30S pre-rRNAs ( Fig . 2d and Fig . S2 ) . These findings suggest that the BMS1 p . R930H mutation results in a reduced activity of BMS1 during the processing of 18S pre-rRNAs . Mutations in several genes that affect ribosomal function have been described to result in a G1/S phase transition delay that may involve p53-dependent and p53-independent pathways , presumably due to “nucleolar stress” resulting from increased free ribosomal proteins as a consequence of ribosomal processing or assembly abnormalities [11] . To determine whether ACC fibroblasts that carry the BMS1 p . R930H mutation show a similar cellular response as in some ribosomopathies , these fibroblasts were examined for their proliferation rate and whether a G1/S phase transition delay can be observed . Skin fibroblasts were isolated from an affected family member through a skin biopsy and expanded in vitro . Control fibroblasts were obtained from an unrelated unaffected individual , matched for ethnicity , age and anatomic location . Subconfluent early-passage ACC fibroblasts and control fibroblasts were analyzed for their cell cycle status , cell proliferation rate and cell migration ability . FACS-based cell cycle analysis showed a G1/S phase transition defect in ACC fibroblasts ( Fig . 3a ) , with a significantly reduced cell proliferation rate ( Fig . 3b ) . Furthermore , cell migration rate was increased in ACC fibroblasts in an in vitro scratch assay , which is consistent with the observation that cell migration is favored in G1 phase ( Fig . 3c ) [12] . Thus , similarly as has been observed in some ribosomopathies , ACC fibroblasts with the BMS1 p . R930H mutation show a link between a mutation in a gene involved in ribosomal function and a G1/S phase transition delay that results in a reduced cell proliferation rate . To determine the molecular changes that are associated with a G1/S phase transition defect and a reduced cell proliferation rate in ACC , unbiased analyses of primary ACCBMS1 ( p . R930H ) fibroblasts compared to control fibroblasts were performed using global gene expression profiling and quantitative comparative proteomic experiments . Microarray gene expression experiments showed differential expression of 459 genes ( p-value<0 . 05 , FDR<5% ) between ACCBMS1 ( p . R930H ) fibroblasts and cells obtained from an unaffected unrelated individual ( Table S1 ) . Increased expression of p21 ( CDKN1A ) mRNA in ACC fibroblasts was noticed , which was further confirmed by semiquantitative RT-PCR ( Fig . 4a and 4b ) . The cyclin-dependent kinase inhibitor p21 mediates a G1/S phase arrest in response to cellular stress and other stimuli , and the increased p21 mRNA levels are consistent with the observed G1/S phase transition delay in ACC fibroblasts . Western blotting experiments confirmed increased p21 protein levels in ACC fibroblasts compared to control fibroblasts , whereas total p53 protein levels were not significantly increased ( Fig . 4d ) . There was also no difference in p53Lys342 acetylation and no p53Ser15 phosphorylation was detected in ACC fibroblasts ( data not shown ) . These findings suggest a role for increased p21 levels in the observed G1/S phase transition defect in ACC fibroblasts . Serine/arginine-rich splicing factor 3 ( SRSF3 ) was found to be downregulated in ACC fibroblasts ( Fig . 4a and 4b ) , which has recently been reported to promote transcription of G1/S phase checkpoint regulators and silencing of SRSF3 caused a G1/S phase arrest ( Kurokawa et al . , EMBO meeting 2011 abstract ) . Overexpression of the mutant BMS1 p . R930H in wild-type fibroblasts resulted in increased transcript levels of p21 and decreased SRSF3 transcript levels , as observed in ACCBMS1 ( p . R930H ) fibroblasts ( Fig . 4c ) . Next , a global quantitative comparative proteomic analysis was performed using iTRAQ-labeling and subsequent MS analysis , comparing control versus ACC early passage subconfluent fibroblasts . Using stringent parameters for statistical significance , 25 proteins were found to be consistently upregulated in ACC fibroblasts , whereas 18 were downregulated ( Fig . 5a ) . Downregulated proteins in ACC fibroblasts included several serine/arginine-rich splicing factors ( SRSF1 , SRSF2 , SRSF3 , SRSF7 ) , including SRSF3 that was also downregulated in the microarray transcript profiling experiments . Other downregulated proteins included heterogenous nuclear ribonucleoproteins ( HNRNPA2B1 , HNRNPH2 , HNRNPA1 ) . Decreased levels of hnRNPA2B1 in ACC fibroblasts were further confirmed by Western blotting experiments ( Fig . 5c ) . Importantly , hnRNPA2 knockdown has been shown to result in a p53-independent increase of p21 levels and an inhibition of cell proliferation [13] , as observed here in ACCBMS1 ( p . R930H ) fibroblasts . Functional enrichment analysis of the proteomic data ranked the category “RNA post-transcriptional modification” as the top-ranked cellular process altered in ACC fibroblasts ( Benjamini-Hochberg multiple test corrected p-value 4 . 72E-06 ) . Network analysis identified the two most significantly altered networks to include top functions for “RNA-posttranscriptional modification” and “cell cycle” ( Fisher exact test –lg p-value of 70 and 21 respectively ) ( Fig . 5b ) . Thus , the results of the unbiased global proteomic analysis in ACCBMS1 ( p . R930H ) fibroblasts are consistent with a defect in pre-rRNA processing . These two top-ranked networks were merged to generate a combined network of the differentially present proteins in ACCBMS1 ( p . R930H ) fibroblasts ( Fig . 6a ) , which were further analyzed for interactions with the differentially expressed transcripts . This analysis revealed the highest number of interactions to include p21 ( CDKN1A ) and hnRNPA2B1 ( Fig . 6a ) . Furthermore , interaction analysis among p21 and the entire merged proteomic network revealed a central role of p21 , suggesting that p21 activation is a central determinant of the ACC phenotype ( Fig . 6b ) .
Mutations in genes for structural proteins of the ribosome or in other genes involved in ribosome biogenesis or function have been found in rare congenital diseases termed ribosomopathies , including Diamond Blackfan anemia ( e . g . RPS19 and RPS24 ) , Shwachman-Diamond syndrome ( SBDS ) , X-linked dyskeratosis congenita ( DKC1 ) , cartilage hair hypoplasia ( RMRP ) and Treacher Collins syndrome ( TCOF1 , POLR1C and POLR1D ) [14] , [15] , [16] , [17] , [18] . However , a disease phenotype in these disorders may not always be a direct consequence of ribosomal dysfunction , but may be due to other disease mechanisms . For example , in both dyskeratosis congenita and cartilage hair hypoplasia ribonucleoprotein complexes containing telomerase are involved , likely critical to maintain stem cell function . RMRP , which is mutated in cartilage hair hypoplasia , interacts with TERT ( the catalytic subunit of telomerase ) and forms a ribonucleoprotein complex that has RNA-dependent RNA polymerase activity [19] . DKC1 , mutated in X-linked dyskeratosis congenita , is associated with small nucleolar RNAs , but also with human telomerase RNA ( TERC ) , and it has been suggested that the disease phenotype in dyskeratosis congenita results from a defect in telomere maintenance [20] . Thus , although these two diseases have been classified as ribosomopathies , their pathologies may result from cellular dysfunction that is not due to a ribosome biogenesis defect . Notably , ribosomopathies manifest with very specific clinical features affecting only few organ systems and often resulting in hematologic and craniofacial abnormalities [21] . These specific phenotypes affect particular cell types , despite the ubiquitous expression of the mutated gene , and illustrate that a mutation in a gene involved in ribosomal function that affects basic cellular pathways can manifest with a selected clinical abnormality . In this context , it is not surprising that despite the ubiquitous expression of BMS1 , individuals harboring the BMS1 p . R930H mutation in non-syndromic ACC display only a localized skin morphogenesis defect without further systemic anomalies . A mutation that slows cell proliferation may manifest itself at an anatomic location that requires rapid growth of a tissue compartment , such as the embryonic skin at the vertex area during the rapid expansion of the skull in embryonic development . As such , it is also not surprising that aplasia cutis can be seen in various diverse syndromes in which cell proliferation rate is affected in addition to other basic cellular pathways that also affect other organ systems and lead to multiple congenital anomalies . For example , Adams-Oliver syndrome ( AOS ) patients have aplasia cutis and multiple additional congenital anomalies , and skin fibroblasts from AOS families with a mutation in a Cdc42/Rac1 regulatory protein ( ARHGAP31 ) showed an enhanced cell migration rate and a reduced cell proliferation rate in vitro , as seen here in ACCBMS1 ( p . R930H ) fibroblasts [3] . Several studies have shown that ribosomal gene mutations can lead to “nucleolar stress” and a G1/S phase cell cycle arrest both via p53-dependent and p53-independent mechanisms that are only partially understood [21] , [22] . It has been proposed that the ribosomal stress response , as a consequence of alterations in ribosomal assembly or processing , results in increased free ribosomal proteins that can either inactivate Mdm2 , resulting in p53 accumulation and p21-mediated cell cycle arrest , or increase p27 levels that ultimately result in a cell cycle arrest [11] . Consistent with these findings in some ribosomopathies , ACC fibroblasts with the BMS1 p . R930H mutation showed a G1/S phase cell cycle transition defect associated with increased p21-levels that result in a reduced cell proliferation rate . BMS1 function has been studied mostly in yeast , demonstrating that inducible depletion of BMS1 in yeast results in inhibition of pre-rRNA processing at sites A0 , A1 and A2 affecting the small ribosomal subunit processome [7] , [8] , [9] , [10] , while little is known about the function of BMS1 in vertebrates . The data presented here are consistent with the findings in yeast and suggest that reduced BMS1 function delays maturation of the 18S rRNA in human cells . A homozygous mutation in the GTPase domain of BMS1 has recently been reported to result in impaired liver development in zebrafish , while heterozygosity for this mutation does not cause an observable phenotype [23] . This observation suggests a role for BMS1 and other components of the small subunit processome in liver development , which is further supported by the identification of homozygous mutations in the gene cirhin in North American Indian childhood cirrhosis that is required for proper 18S rRNA maturation [24] , [25] . The heterozygous BMS1p . R930H mutation in ACC is located in the putative regulatory GAP domain , and therefore is likely to affect BMS1 function to a lesser degree than the homozygous mutation in the GTPase domain of BMS1 in zebrafish . In this context , it is not surprising that patients with autosomal dominant ACC do not display major developmental liver abnormalities . Instead , the data presented here show a developmental localized skin formation defect at a site of rapid expansion of the skin due to a heterozygous mutation in the putative GAP domain of BMS1 , which provides the first human mutation for BMS1 . Through which exact mechanisms this mutation causes a p21-mediated G1/S phase arrest and is correlated with the observed downregulation of hnRNPs and SRSFs remains to be determined in future studies . The reported observation that hnRNPA2 knockdown results in a p53-independent increase of p21 levels and an inhibition of cell proliferation [13] , similar as observed here in ACCBMS1 ( p . R930H ) fibroblasts , together with the results form the interaction analyses of the proteomic data that show the largest number of interactions to involve p21 and hnRNPA2B1 in ACC fibroblasts , suggest that the reduced protein levels of hnRNPA2B1 likely play a role in ACC pathogenesis . During embryonic development p21 expression correlates with arrest of cell proliferation and is found in postmitotic cells immediately adjacent to the proliferative compartment , as in the outer embryonic epidermis and particularly in the developing hair follicles [26] , [27] . Notably , p21 expression is decreased in terminally differentiated cells and overexpression of p21 inhibits late stages of differentiation of keratinocytes and the stem-cell potential of keratinocyte subpopulations [28] , [29] . Thus , the increased p21 levels in ACC may result in an inhibition of cell proliferation as well as an inhibition of terminal differentiation of the outer epidermis during embryonic development , and explain the observed skin morphogenesis defect in ACC . Increased p21 levels have also been linked to increased scar formation , which is consistent with the prominent hypertrophic scar formation in patients with ACC , as was also observed in affected members of this family ( Fig . 1a ) [30] . In summary , the findings in this study show that the BMS1 p . R930H mutation is associated with a downregulation of hnRNPA2B1 ( and other hnRNPs and SRSFs ) and a p21 upregulation , leading to a G1/S cell cycle phase transition delay and an inhibition of cell proliferation . The data presented here provide a novel link between BMS1 , a p21-mediated cell cycle arrest and skin morphogenesis .
A five-generation family was identified with autosomal dominant inheritance of ACC . All participants provided written consent , and the Institutional Review Board of Massachusetts General Hospital approved this study . Genomic DNA was extracted from peripheral blood lymphocytes using the QIAGEN Puregene blood isolation kit ( Qiagen ) . DNA concentrations were determined using the Picogreen assay ( Life Technologies ) . Genomic DNA from individuals of this family were prepared , labeled and hybridized to the Affymetrix Genome-Wide Human SNP array 6 . 0 , which features more than 906 , 000 SNPs . SNP data was analyzed with the Affymetrix Genotyping Console . Merlin version 1 . 1 . 2 was used for parametric linkage analysis [31] . Linkage analysis was performed with either 100% or 95% disease penetrance , equal or calculated allele frequencies , and various disease allele frequencies . The frequency of ACC in the population has been estimated to be about 1∶30 , 000 births . Independently , microsatellite genotyping was performed using informative markers from A&B Biosciences that spanned chromosome 10 . About 10 µg of DNA from one affected individual of this family was used for exome capture using the Agilent Sure select 50 Mb kit . Sequencing was performed on an Illumina HiSeq 2000 sequencer . Reads were mapped to the UCSC hg19 reference human genome . Sequence data was analyzed and variants were filtered using the DNAnexus software package . Bidirectional Sanger sequencing of PCR amplicons from genomic DNA was used to confirm the presence and identity of variants identified via exome sequencing . Primary fibroblasts were obtained from a family member with ACC ( p . R930H ) through a 4 mm abdominal skin biopsy . Control fibroblasts were obtained from a skin biopsy of a matched unrelated individual without ACC ( matched for anatomic location , skin phototype , and age ) . The skin biopsy sample was treated with collagenase I , and subsequently fibroblasts were maintained in culture in DMEM ( Invitrogen ) with 20% fetal calf serum ( Sigma ) and antibiotic/antimycotic ( Invitrogen ) . For cell immunolabeling experiments , fibroblasts were grown on poly-L-lysine coated glass slides ( BD Biosciences ) and fixed with methanol . Primary antibodies used were as follows: polyclonal rabbit BMS1 ( Sigma ) , monoclonal mouse BMS1 ( Santa Cruz ) , rabbit polyclonal nucleophosmin ( Invitrogen ) , rabbit polyclonal anti-phospho Histone H3 ( Ser10 ) ( Millipore ) . Secondary antibodies used were fluorescently labeled Alexa antibodies from Invitrogen . Nuclei were stained with DAPI . Mouse embryos at E13 . 5 were fixed in 4% PFA and embedded in OCT and cryosections were used for immunofluorescence experiments . F-actin labeling was performed with phalloidin-Alexa Fluor 488 conjugates ( Invitrogen ) . Microscopy was performed with a Zeiss Axiovert microscope . In vitro mutagenesis was performed according to the manufacturers protocol ( Stratagene ) . Mutant ( c . 2789G>A ) and wild-type BMS1 cDNA was cloned into the pEGFPN1 ( Clontech ) expression vector , fusing EGFP at the C-terminus of the expressed cDNA . Primary human wild-type fibroblasts that were used as controls in all experiments ( matched for age- , anatomic location and skin phototype ) were transfected in triplicate with purified plasmid using XtremeGene HP DNA transfection reagent ( Roche ) or nucleofection using the Amaxa human dermal fibroblast nucleofector kit with program U-023 ( Amaxa ) . Expression and subcellular localization of the expressed EGFP-tagged protein was evaluated on a Zeiss Axiovert fluorescence microscope . For semiquantitative RT-PCR experiments , transfected fibroblasts ( n = 3 independent transfections ) were sorted for EGFP 48 hours after transfection using a cell sorter ( MoFlo ) and RNA was immediately isolated after sorting with Trizol reagent ( Invitrogen ) . Overexpression was confirmed by semiquantitative RT-PCR for BMS1 transcript levels . All experiments were performed in triplicate . Cell cycle analysis with propidium iodide was performed according to standard protocols . Briefly , subconfluent fibroblasts were fixed in 70% ethanol , treated with RNase A and stained with propidium iodide ( Sigma ) and cell cycle status was determined by FACS analysis and subsequent analysis by FlowJo software ( version 9 . 4 . 9 ) using the Watson pragmatic model for determination of cell cycle phases . Experiments were performed in triplicate for ACC and WT cells . The CyQUANT fluorescence-based microplate assay was used for quantitation of cell numbers and used according to the manufacturers protocol ( Invitrogen ) . Binding to cellular nucleic acids was measured by using 485 nm excitation and 530 nm emission filters with a fluorescence microplate reader . The fluorescence emission of the dye-nucleic acid complexes was then correlated linearly with cell numbers from a dilution series of cells that were collected d2–d7 after cell culture of equal numbers of cells . For each time point triplicate experiments were performed . Cell migration assays were performed with WT and ACCBMS1 ( p . R930H ) primary dermal fibroblasts plated on fibronectin-coated 6 cm2 tissue culture dishes ( n = 3/group ) . Cells were serum- and growth-supplement- starved for 12 hours before inducing a linear scratch wound with a 1 ml pipette tip . Cultures were washed twice with PBS , and wound margins were photographed . The same wound margin fields were photographed after 14 hours , and the scratch areas that were not repopulated by fibroblasts were measured with Zeiss Axiovision software . Similar findings were obtained when dishes were coated with collagen . Western blotting was performed with 50 µg of total protein lysate from subconfluent early passage ACC and control fibroblasts . Protein concentrations were determined by a standard Bradford assay . NP-40 lysis buffer and complete protease inhibitor cocktail ( Roche ) was used . NuPage Bis-Tris 4–12% gradient gels were used . Primary antibodies used were: mouse monoclonal anti-p53 ( clone DO-1 ) ( Becton Dickson ) , rabbit polyclonal anti-β-actin ( Thermo ) , rabbit anti-p21 polyclonal antibody ( Thermo Fisher ) and mouse anti-HNRNPA2B1 ( Millipore ) . HRP-conjugated secondary antibodies were used from GE . β-actin loading control was performed on each blot . Band intensities were quantified using NIH ImageJ software ( NIH , version 1 . 46 ) . To study the kinetics of pre-rRNA processing pulse-chase labeling with L-[methyl-3H]methionine was performed on early passage subconfluent ACCBMS1 ( p . R930H ) fibroblasts and control fibroblasts as described [32] . Briefly , cells were cultured on 6-well plates and after being cultured for 15 minutes in methionine-free medium , cells were pulsed with L-[methyl-3H]methionine at a final concentration of 50 µCi/ml for 30 minutes . 10× methionine medium was used for the chase and total RNA was isolated with Trizol ( Invitrogen ) at 0 , 15 , 45 , 60 and 180 minutes after the end of the pulse . 1 µg of total RNA was loaded on a 1% formaldehyde-agarose gel and after electrophoresis blotted overnight on a nylon membrane ( Ambion ) . Signal intensity was increased with the EN3HANCE spray ( PerkinElmer Life ) and treatment with carbon tetrachloride ( Sigma ) as described [32] . The nylon membrane was exposed to Kodak BioMax MS high-sensitivity radiography film ( Kodak ) . Early-passage fibroblasts isolated from an affected individual with ACC from this family with the BMS1 mutation p . R930H were used for comparative proteomic analysis . Fibroblasts , matched for ethnicity and anatomic location , were used as control . From each individual fibroblasts were used as quadruplicate experimental samples in a 8-plex iTRAQ experiment . Whole cell lysates were generated using Pressure Cycling technology . Collected cells were resuspended in RIPA buffer plus Pefablock ( Roche complete tabs , 1 tablet per 10 ml ) and 1 mM DTT . The cells were then lysed in a Barocyler NEP2320 instrument from Pressure Biosciences , Inc . ( South Easton , MA ) . Following lysis , proteins were recovered by acetone precipitation . Acetone precipitated material that resuspended in milliQ water was used as the soluble protein extract . Protein concentrations were determined using the Bio-Rad Protein Assay with BSA used to generate a standard curve . The 8-plex iTRAQ assay for multiplexed relative quantitation ( AB SCIEX , Foster City , CA ) was used to determine the protein level differences between control and ACC cells . Briefly , equal amounts of soluble protein from each fibroblast experimental group ( 100 µg ) was dried in a speed vacuum/centrifuge . Following the recommended 8-plex iTRAQ protocol , dried proteins were resuspended in 21 µl of 500 mM TEAB ( pH 8 . 5 ) , 0 . 1% SDS . Proteins were then reduced with 5 mM TCEP by incubation at 55°C for 1 hour . Reduced disulfide bonds were then blocked by adding a final concentration of 8 mM MMTS and incubating at room temperature for 10 min . Peptides were generated in each sample by overnight digestion with trypsin ( AB SCIEX ) added at a ratio of 1∶10 . Samples were then labeled with the 8-plex iTRAQ reagent . Confirmation of efficient label attachment was confirmed by MS/MS in an AB SCIEX 4800 MALDI-TOF/TOF before samples were pooled into one . Half of the sample ( ∼400 µg ) was processed using standard MuDPIT methodology to maximize protein identification . First dimension strong cation exchange ( SCX ) separation was performed on an Agilent 1100/1200 HPLC with a POROS HS/20 column ( 4 . 6×100 mm , AB SCIEX ) . 40 fractions covering the eluted peptides were then individually run for second dimension reverse phase separation on a Dionex Ultimate Plus nanoLC with a C18 Acclaim PepMap100 column ( 75 µm×15 cm , 3 µm beads ) . Approximately 500 spots were printed per fraction to 4800 MALDI plates ( 5 SCX fractions per plate ) with CHCA ( 5 . 0 mg/ml ) mixed in by the ProBot in a mixing Tee immediately before printing . MS and MS/MS on the peptides were performed in the AB SCIEX 4800 mass spectrometer . Relative abundance quantitation and peptide and protein identification were performed using ProteinPilot software 3 . 0 ( Applied Biosystems ) . The Swiss-Prot Homo sapiens protein database ( Swiss-Prot/UniProt ) was used for all searches . The data were normalized for loading error by bias analysis using ProteinPilot . The high confidence proteins are defined as those with ( a ) >90% confidence as determined by ProteinPilot ( ProtScore ≥1 . 0 ) and ( b ) two distinct peptides with different iTRAQ spectra identified with ≥95% confidence . These high confidence proteins were identified using Perl and R scripts . The differentially present proteins were identified by comparing the relative weighted average iTRAQ values of proteins in the control and ACC groups . The significance of differential levels of protein iTRAQ ratios was determined using the multiple test corrected ( FDR ) P-value . Proteins that had a iTRAQ ratio >1 . 2 or <0 . 80 ( calculated on basis of variance in iTRAQ values ) , a raw P-value of <0 . 05 and a FDR<10% between control and ACC groups were considered differentially present . Interactive networks and pathways were analyzed using the Ingenuity Pathways Analysis package ( IPA 4 . 0 ) ( http://www . ingenuity . com/ ) . It calculates the P-value for each network , and function according to the fit of user's data to IPA database . It displays the results as Fisher's exact test -lg P-value indicating the likelihood of a gene to be found in a network or function by random chance . Early-passage fibroblasts isolated from an affected individual with aplasia cutis from this family with the BMS1 mutation p . R930H and control fibroblasts , matched for ethnicity and anatomic location ( but not for gender ) , were each used as triplicate experimental samples in gene expression analysis experiments ( which were also used in the comparative proteomic analysis ) . Total RNA was isolated from fibroblasts using the RNEasy Plus kit ( Qiagen ) , and treated with DNAse . Semiquantitative RT-PCR was performed using the total RNA . cDNA was generated using the Transcriptor First Strand cDNA synthesis kit with random hexamer primers ( Roche ) . Semiquantitative RT-PCR was performed using the LightCycler 480 DNA SYBR-Green I Master kit on a LightCycler 480 ( Roche Applied Science ) . The following primers were used: p21 ( up: GCAGACCAGCATGACAGATTT; dw: GGATTAGGGCTTCCTCTTGGA ) . Pretested SRSF3 and BMS1 qPCR primers were used from SABiosciences ( Qiagen ) and results were normalized to 36B4 transcript levels . Reactions were performed in triplicate ( n = 3 ) and also three times for each sample . Human Genome U133 Plus 2 . 0 Affymetrix microarrays were used for global gene expression profiling experiments . Obtained raw data was analyzed using the dCHIP software ( https://sites . google . com/site/dchipsoft/ ) . Differential expression was determined significant with dCHIP with a P-value<0 . 05 and a false discovery rate ( FDR ) <5% among the control ( WT ) and ACC groups ( samples in triplicate for each group; corrected for gender ) . To achieve a stable partial inducible knockdown of BMS1 a puromycin-selectable TRIPZ doxycycline-inducible lentiviral shRNA system was used , which allows for the detection of shRNA-mediated gene silencing by TurboRFP expression ( Open Biosystems ) . TRIPZ BMS1 shRNA clone V3THS_361658 was used to transfect wild-type fibroblasts and cells were selected by puromycin treatment . Fluorescence microscopy confirmed that after puromycin treatment all remaining cells were expressing TurboRFP . After 48 hours treatment with 1 µg/ml doxycycline RNA was isolated from these stably transfected cells to obtain cDNA . Semiquantitative RT-PCR for BMS1 confirmed a ∼50–60% knockdown of BMS1 transcript levels and cDNA was used for semiquantitative RT-PCR of p21 and SRSF3 transcript levels . As controls these cells were used in the absence of doxycycline treatment . From each group experiments were performed in triplicate . Total RNA was used for Northern Blot analyses . Northern blotting was performed as described previously [32] . From each sample 2 µg total RNA was loaded on a 1 . 2% formaldehyde-agarose gel and blotted on a positively charged Nylon membrane ( BrightStar-Plus , Ambion ) . Radioactive probes labeled with [γ-32P]ATP for ITS-1 , ETS-1 and ITS-2 were used as described previously [33] . Band intensities were quantified using NIH ImageJ software ( NIH , version 1 . 46 ) .
|
Elucidating the pathomechanisms in congenital diseases of the skin provides the opportunity to learn what cellular processes are important during embryonic development of the skin structures . Aplasia cutis congenita ( ACC ) manifests with localized skin defects , most commonly affecting the scalp skin . Here , global proteomic and transcriptional analyses are combined with genome-wide linkage and exome sequencing approaches to identify the molecular mechanisms involved in ACC . A mutation in the ribosomal GTPase BMS1 is identified in ACC that affects 18S rRNA maturation . This mutation is associated with a p21-mediated G1/S phase transition delay during the cell cycle that inhibits cell proliferation . The findings are consistent with mutations in ribosomal disorders that result in nucleolar stress and a G1/S phase transition delay . Thus , mutations in BMS1 can affect the formation of a highly proliferative tissue during development , such as the rapidly expanding scalp epidermis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"dermatology",
"genetics",
"biology",
"human",
"genetics",
"genetics",
"of",
"disease"
] |
2013
|
BMS1 Is Mutated in Aplasia Cutis Congenita
|
Because physical form and function are intimately linked , mechanisms that maintain cell shape and size within strict limits are likely to be important for a wide variety of biological processes . However , while intrinsic controls have been found to contribute to the relatively well-defined shape of bacteria and yeast cells , the extent to which individual cells from a multicellular animal control their plastic form remains unclear . Here , using micropatterned lines to limit cell extension to one dimension , we show that cells spread to a characteristic steady-state length that is independent of cell size , pattern width , and cortical actin . Instead , homeostatic length control on lines depends on a population of dynamic microtubules that lead during cell extension , and that are aligned along the long cell axis as the result of interactions of microtubule plus ends with the lateral cell cortex . Similarly , during the development of the zebrafish neural tube , elongated neuroepithelial cells maintain a relatively well-defined length that is independent of cell size but dependent upon oriented microtubules . A simple , quantitative model of cellular extension driven by microtubules recapitulates cell elongation on lines , the steady-state distribution of microtubules , and cell length homeostasis , and predicts the effects of microtubule inhibitors on cell length . Together this experimental and theoretical analysis suggests that microtubule dynamics impose unexpected limits on cell geometry that enable cells to regulate their length . Since cells are the building blocks and architects of tissue morphogenesis , such intrinsically defined limits may be important for development and homeostasis in multicellular organisms .
The physical properties of a system depend to a large extent upon its scale . Therefore , it is not surprising to find that many biological structures are maintained within relatively tightly constrained size limits [1] , [2] . In some cases , the dimensions of macromolecular assemblies are enforced by “molecular rulers” like titin , which helps to govern the length of the sarcomeric repeats in muscle [3] . However , many seemingly stable structures , such as metaphase spindles [4] and cilia [1] , exist in a state of dynamic equilibrium in which a stable form arises from the collective action of a large number of molecular machines functioning in concert . Although mechanisms have been proposed for the control of the length of such polymers [1] , through for example length-dependent microtubule depolymerisation [5] , little is known about this fundamental and widespread biological phenomenon . For unicellular organisms , intrinsic mechanisms have been identified that regulate cell shape [2] , [6] , maintain a steady-state cell size , and couple cell length and size [7] . However , it remains unclear whether similar controls regulate the dimensions of cells from multicellular animals , which , by virtue of not having a cell wall , assume a form that is plastic and a variable size , both of which depend to a large degree upon the extracellular tissue environment in which cells find themselves [8] , [9] . Nevertheless , since form and function are intimately linked and vary from cell type to cell type , it seems likely that the shape of many animal cells will be maintained within intrinsically defined limits . Such behaviour has been observed in assays of cell spreading [10] and cell migration on planar adhesive substrates [11] , [12] . Moreover , studies of cells on grooved , scratched , or patterned substrates have in some cases [13] , [14] revealed limits to cell extension . In addition , regulated changes in cell geometry have long been known to drive a variety of morphogenesis movements in developing animals . During Drosophila development , for example , changes in epithelial cell shape and height are thought to drive internalisation of the ventral furrow [15] . Similarly , during neural tube development in zebrafish , individual neuroepithelial cells act together to form a double-layered epithelium with a defined width , even under conditions in which the entire structure is mis-positioned or duplicated [16] . Here , to systematically investigate the mechanisms that control animal cell geometry , we employed micro-contact printing [17] to generate adhesive lines of extracellular matrix that limit cell width but leave cells free to regulate their geometry in the other two dimensions ( length and height ) . This analysis of cells spreading on adhesive lines reveals that animal cells can spread to a characteristic steady-state length that is independent of cell size , pattern width , and cortical actin , but is dependent on a population of microtubules that aligns along the long axis of cells as the result of interactions between microtubule plus ends and the cell cortex . Similarly , a population of oriented microtubules mediates length control in epithelial cells of the developing zebrafish neural tube . A mathematical model shows that cell length homeostasis in these cases can be quantitatively explained by the collective action of dynamic , orientated microtubules as they drive cell extension and undergo cortex-dependent catastrophe . Together , this experimental and theoretical analysis reveals a role for microtubules in homeostatic cell length control in HeLa cells , Drosophila S2R+ cells , and zebrafish neuroepithelial cells , suggesting that it may be an important general feature of animal cell biology .
To explore the intrinsic regulation of cell shape , we began by seeding a population of freshly harvested exponentially growing HeLa cells onto micropatterned fibronectin lines ( Figure 1A ) ranging in width from 3 to 35 µm , separated by non-adhesive polyethylene glycol , and onto equivalent non-patterned areas of the substrate ( Figure 1B ) . Other researchers previously performed similar experiments by plating cells on grooved or scratched substrates [18]–[20] or on adhesive strips [13] . In our case , after allowing 2 h for cells to adhere to and spread on the micropatterned substrate , cell length was monitored using semi-automated software designed to remove user bias and to facilitate the analysis of large datasets ( Figure S1 ) . Cell length for this analysis was defined as the maximum distance , parallel to the patterned line , separating extensions at distal cell tips ( Figure 1A ) . Unexpectedly , cells from an exponentially growing population ( with a wide range of masses ) spread to a relatively well-defined average length of 44±10 µm , which proved largely independent of line width and similar to that of cells on non-patterned substrates ( ∞ ) ( Figure 1C ) . Cell spreading in this assay was accompanied by a corresponding change in cell height , as expected if cell volume is conserved ( Figure 1E and data not shown ) . These observations were similar to those previously reported for fibroblasts on scratched substrates [13] . Because a fixed-time-point assay was used for this analysis , however , the independence of cell length and pattern width could be the result of either a constant rate of cell spreading or the action of a cell length control mechanism . When live-cell imaging was used to examine the kinetics of cell spreading , we observed HeLa cells spreading monotonically over a period of approximately 60 min , reaching a steady-state length that was independent of pattern width ( Figure 1D and 1E ) . These data suggest that HeLa cells have an intrinsically defined length . To determine whether cell length control is a peculiarity of HeLa cells spreading on a fibronectin substrate or a more general feature of animal cell biology , we repeated these experiments using the adherent Drosophila hemocyte-derived S2R+ cells [21] , which are significantly smaller than HeLa cells ( with a mean volume , measured using an automated cell counter , of 1 , 177±64 µm3 for Drosophila S2R+ cells compared to 2 , 121±1 , 116 µm3 for HeLa cells; Figure 2A ) . These cells do not adhere well to fibronection-coated substrates , but could be induced to spread on glass dishes coated with Concanavalin A ( ConA ) [15] . Once spreading was complete ( 5 h after plating ) , the length of these cells was measured . Like HeLa cells , patterned S2R+ cells were found to achieve a reproducible steady-state length that was relatively independent of their width ( Figure 2B and data not shown ) . Moreover , despite their very different average volumes , S2R+ and HeLa cells had comparable resting lengths ( 55±9 µm and 44±10 µm , respectively ) . This observation suggested the possibility that the limit of an animal cell's long axis might be regulated independently of its volume . As a direct test of this hypothesis we altered the culturing conditions to obtain a population of HeLa cells grown to confluence that had an average volume of 1 , 233±835 µm3 ( about half that of HeLa cells grown to 50% confluence; Figure 2A ) . When the two populations of differently sized HeLa cells were plated onto micro-contact-printed fibronectin lines of varying widths for 2 h , they were found to spread to statistically similar lengths ( Figure 2B ) despite having different heights ( data not shown ) , confirming the independence of cell length and volume . It seemed unlikely that cell length control depends on plasma membrane tension , since dramatic changes in cell volume and line width had little impact on the length of HeLa or S2R+ cells , or on the spreading kinetics of HeLa cells on narrow and thick patterned lines ( Figure 1D ) . In searching for mechanisms underlying cell length control we therefore turned to examine a possible role for the cytoskeleton [22] , since actin filaments and microtubules have been implicated in the control of fixed-length structures [1] , [3] , [23] and in the regulation of cell length [6] , [13] and cell spreading [18] , [24]–[27] . We began by analysing the cytoskeletal changes that accompany cell spreading on micropatterned lines . Following attachment to a patterned surface and a brief period of blebbing ( Figure 1E ) , cells developed spreading lamellipodia that quickly reached the edges of the patterned lines . While lamellipodia reaching laterally outside of the pattern into regions passified with polyethylene glycol underwent periodic cycles of extension and retraction [28] , lamellipodia extending in the direction of cell elongation remained tightly bound to the adhesive substrate ( data not shown ) . To test whether these actin-based lamellipodia play a role in cell elongation or cell length control , we took advantage of the efficacy of RNA interference ( RNAi ) –mediated gene silencing in fly cell culture to target SCAR/WAVE ( hereafter SCAR ) to selectively remove these structures from spreading cells [29] . Five days after treatment with control ( LacZ ) or SCAR double-stranded RNA ( dsRNA ) [30] , [31] , Drosophila S2R+ cells were seeded onto non-patterned substrates ( Figure 3A and 3B ) and onto micropatterned ConA lines ( Figure 3C ) . Cell length was then assayed 5 h later ( Figure 3D ) . As Figure 3C shows , SCAR RNAi cells assumed their typical RNAi phenotype , in which lamellipodia are replaced by long , radial , microtubule-rich processes [31] . Despite this , the majority of microtubule-based protrusions became oriented along the ConA lines . When we measured the distance parallel to the pattern between distal protrusion tips in these cells , we found that the length of SCAR RNAi cells was statistically indistinguishable from that of control cells ( Figure 3D ) . Similarly , control RNAi and SCAR RNAi cells reached similar steady-state spread diameters on non-patterned substrates [31] ( Figure 3D ) . Thus , although SCAR is required for the formation of lamellipodial-based protrusions , lamellipodial actin is not required for cell elongation and does not alter the steady-state length of microtubule-rich extensions . We then repeated these experiments in HeLa cells , using a Rac inhibitor [32] to compromise lamellipodial formation ( Figure 3F ) . As with S2R+ cells , this did not alter cell length ( compare Figure 3F and 3G ) . To test a possible role for actin-myosin-mediated cortical tension in the regulation of cell length , we also carried out a similar analysis using blebbistatin to inhibit Myosin II [33] . Once again , this did not affect the rate of cell spreading ( Figure 3H and 3I ) or resting cell length ( 54 . 3 [−8 , +17] µm for HeLa cells [n = 345] on fibronectin lines , and 50 . 9 [−4 , +11] µm for HeLa cells [n = 52] on non-patterned fibronectin ) ( Figure 3E ) . Although surprising , these data concur with previously reported work showing that the final spread area of fibroblasts and HeLa cells is relatively independent of membrane tension [34] and cortical actin [10] . Having failed to identify a role for the actin cytoskeleton , these data prompted us to examine the function of microtubules , which have previously been implicated in cell spreading [20] , [24]–[26] , [35] and in the generation of an elongated cell shape [36] , [37] . Strikingly , in the presence of the microtubule inhibitor colcemid [38] , HeLa cells were unable to elongate ( Figure 4A and 4D ) . This effect was reversible since cells re-spread to their characteristic length after the drug was washed out and the microtubule cytoskeleton re-established ( Figure 4C , 4E , and 4F ) . Cell spreading in this system therefore requires microtubules . To gain mechanistic insight into their precise role , we then analysed microtubule organisation at intervals during cell spreading ( Figure 5 ) . Cell elongation was accompanied by a progressive polarisation of the microtubule cytoskeleton , as microtubules concentrated on the basal part of the cell ( Figure S3 ) became aligned along the long cell axis ( Figure 5B and 5D; quantified in Figure 5F ) . Strikingly , a similar re-alignment was observed over shorter time scales as individual growing microtubules in cells at steady state became oriented to lie along the long cell axis ( Figure 6A ) , as the result of contacts between growing microtubule plus tips ( marked with EB3-GFP ) and the ruffling cell margin at the interface between an adhesive and non-adhesive substrate ( Figure 6B and 6C ) . This cortex-induced change in microtubule direction was similar to that previously described in yeast [6] , animal cells [20] , [39] , and plant cells [40] . Interestingly , the fate of each growing microtubule meeting the cortex depended strongly on the angle of contact ( Figure 6B , 6C , and 6E ) , such that microtubule plus ends contacting the cortex at a steep angle ( 32°–90° from the long line axis ) underwent catastrophe , while others contacting the cortex at less of an angle ( 0°–25° ) changed their direction of growth to run parallel to the cell edge . This explains the low probability of finding microtubules aligned at an angle of between 30° and 90° with respect to the long cell axis in Figure 5 ( or −30° and −90° according to the notation in Figure 6C ) . As a result of this angular dependency , the microtubule cytoskeleton reached a highly polarised equilibrium state , with the majority of microtubules running along the cell edge ( Figure 6A; quantified in Figure 6F; explained in Figure 6G ) . Importantly , this angular dependency of microtubule catastrophe/bending resulted in overt polarisation of the microtubule cytoskeleton . We also observed oriented microtubules leading during cell extension ( Figure 6D and 6F; Video S1 , S2 , S3 ) . Taken together these data suggest that microtubules oriented parallel to the adhesive boundary are strong candidates for drivers of cell spreading and the source of intrinsic cell length control . To test whether the dynamic behaviour of this oriented array of microtubules could define cell length in this system , we generated a simple mathematical model of microtubule-based cell length control ( Figures 7 and 8 ) . In contrast with previous models of cell spreading [10] , [41] , [42] , our model was constructed to assess the effects of a polarised array of dynamic microtubules on cell elongation and homeostasis . Based upon our analysis of the cellular distribution of microtubules ( Figures 5 and 6 ) , EB3-GFP comets ( Figure 6D ) , and γ-tubulin ( Figure S2 ) , microtubules in the model were assumed to grow out from the cell centre towards cell tips ( Figure 7A ) . Cycles of dynamic instability were then implemented using values of growth and catastrophe rates taken from the experimental literature [43] . The interaction between microtubules and the cell cortex was then modelled by assuming that contact ( i ) increases the microtubule catastrophe rate ( by a factor of 16 [43] ) and ( ii ) drives the extension of the cell margin [44] . ( Although the mechanism by which this occurs is not specified in the model , we think it likely that it is through the delivery of new material required for local growth [24] rather than through force generation [45] , [46] . ) A slow fixed rate of margin retraction was then implemented , based upon measurements of the rate of retraction of cells on lines in the absence of microtubules ( 0 . 4±0 . 3 µm/min ) , which may reflect the turnover of material from the cell periphery . Simulations using this simple scheme were found to recapitulate the path of cell elongation and cell length homeostasis ( Figure 7A–7C ) . Furthermore , the model predicts a linear decrease in microtubule density from the cell centre to the cell edge , which was verified experimentally ( Figure 7D ) . Interestingly , the model also revealed that , irrespective of the actual cell length , it is the small number of dynamic microtubules ( approximately two ) that reach the cell cortex that maintain cell length homeostasis , by countering the tendency of the cell margin to retract ( Figures 7C and S4 ) . This is in line with previous data suggesting that a small population of pioneer microtubules is sufficient in some systems to drive forward movement of the cell edge [47] . When we examined the effects of varying the remaining free parameters in the model ( Figure 8A–8C ) , we found that cell length homeostasis ( the coefficient of variation in cell length ) was marginally sensitive to changes in the value of α , representing the level of microtubule cooperation in the system ( Figure 8A–8C ) , and to changes in Nm , the number of cooperating microtubules . Based on measurements of cell length and microtubule numbers in HeLa cells on lines , we were able to estimate the value of α as 8 , implying that individual microtubules cooperate to drive cell spreading . More significantly , an analysis of the effects of varying the other experimentally determined parameters in the model revealed that cell length control critically requires a high value of cB/cI , the ratio of the microtubule catastrophe rate at the cortex ( cB ) to that in the cell interior ( cI ) . Thus , cell length becomes progressively more variable as the ratio of cB/cI is reduced , e . g . , as cB tends to 0 ( Figures 8D , 8E , and S4B ) . By contrast , changes in the microtubule polymerisation rate ( vg ) induce corresponding changes in cell length in the model without inducing a loss of homeostasis ( Figures S4A , 9A , and 9B ) , causing the system to stabilise at a new steady-state length when the number of microtubule plus ends interacting with the cell cortex returned to the equilibrium value of ∼2 ( Figure 9B ) . This serves as a good test of the likely effects of the addition of a “microtubule inhibitor” on cell length control ( Figure 9A and 9B ) . To test whether this prediction is borne out in experiment , we used an inhibitor of microtubule dynamics [38] to reduce the rate of microtubule polymerisation in HeLa cells on lines of varying width at steady state . After allowing cells 2 h to spread , 40 nM colcemid or an equivalent amount of the carrier DMSO was added to the medium for 30 min ( Figure 9C–9E ) . In the case of colcemid , but not DMSO , this was sufficient to disturb microtubule organisation without causing a complete loss of microtubules ( data not shown ) , and induced active cell shortening ( Figure 9C and 9D ) . As predicted , cells settled down to a new shorter length following this treatment , irrespective of line width ( Figure 9C and 9D ) . We conclude that the dynamic behaviour of the population of longitudinally polarised microtubules plays a key role in homeostatic cell length control . Previous work has suggested roles for microtubules in the regulation of cell shape in 3-D environments [48] . Therefore , we were prompted to test whether cells exhibit a similar type of cell length homeostasis in a tissue and developmental context . We used the zebrafish neural tube as a simple model system for this analysis for several reasons . First , cells in this tissue are bipolar in form and of similar length to S2R+ and HeLa cells on lines . Second , once formed [16] , this tissue is maintained as a stable structure consisting of two parallel columns of highly elongated neuroepithelial cells [49] , making reliable measurements of cell length relatively easy . Third , the tissue is amenable to imaging and perturbation experiments using morpholinos . To begin , we tested whether cell length depends upon cell volume in the zebrafish neural tube by arresting cells in the G2 phase of the cell cycle using an established protocol in which a morpholino against the translational start site of the G2/M regulator Emi1 ( Emi1-MO ) or a control morpholino ( Con-MO ) is injected into the one-cell embryo [50] , [51] . As previously reported , this treatment does not affect cell division until the neural plate stage because of a maternal effect and has very limited cytotoxicity [50] , [51] . For consistency across animals , measurements of cell lengths were then made using the neuroepithelium of the hindbrain close to the developing otic vesicle in 19 somite ( 19s ) –stage embryos . At this stage in development the neural tube has not yet inflated its ventricle , and neuroepithelial cells from the left and right sides meet in the middle of the tube , as confirmed by the expression of the polarity protein Par3-GFP , which reveals the apical ends of all cells lining up along the tube midline in both control and emi1 morphant embryos ( Figure 10A ) . This block in cell cycle progression ( evident in the loss of pH3 staining in Figure 10A ) led to a significant 2 . 6-fold increase in neuroepithelial cell volume ( Figure 10B; Emi1-MO , 9 , 034 . 2±3 , 839 . 3 µm3 , n = 7 cells from three embryos; Con-MO , 3 , 516 . 4±608 . 2 µm3 , n = 8 cells from four embryos; t test , p = 0 . 007 ) when compared to control-morpholino-injected embryos . The variability in the extent of volume increase observed likely reflects the slight variability of the timing with which the Emi1-MO-induced cell cycle arrest kicks in in different embryos . Since the vast majority of individual cells in the tissue span the entire width of the neural tube ( Figure 10A ) , we were able to use the width of the tissue at three locations close to the otic vesicle to estimate average cell length . Despite the large difference in cell volumes between control and emi1 morphant embryos ( readily visible in embryos labelled with Par3-GFP and H2B-RFP; Figure 10A ) , the length of neuroepithelial cells remained unaltered ( Figure 10C; Con-MO , 44 . 97±3 . 66 µm , n = 10; Emi1-MO , 48 . 11±6 . 83 , n = 10; t test , p = 0 . 188 ) . This shows that cell length is independent of cell volume in this tissue context as it is in our cell culture models . To test whether microtubule dynamics are required to define cell length in this system , we visualised microtubule cytoskeletal organisation within the neuroepithelium . Stochastic labelling of cells in the neural tube with GFP fused to the microtubule-associated protein doublecortin ( DCX ) [52] revealed bundles of parallel microtubules running the entire length of each neuroepithelial cell ( Figure 10D ) , from the apical to the basal limit of the epithelium . To determine whether these microtubules function in cell length control , as was shown for microtubules in cells on micropatterned lines , we added low doses of the microtubule inhibitor nocodazole to these embryos for a period of 30 min . While high doses of microtubule inhibitors lead to a reversible loss of neuroepithelial form as all the cells in the tissue round up ( data not shown ) , this treatment leaves overall tissue architecture intact ( Figure 10E ) . The result was a significant reduction in the width of nocodazole-treated neural tubes when compared to those in DMSO control embryos ( Figure 10F ) , without affecting differences in width that characterise different parts of the tissue . Moreover , as the concentration of nocodazole was increased , the neural tube became progressively narrower ( DMSO , 49 . 37±3 . 11 µm , n = 8; 5 µg/ml nocodazole , 48 . 99±2 . 73 µm , n = 9; 10 µg/ml nocodazole , 46 . 93±3 . 29 µm , n = 8; 20 µg/ml nocodazole , 43 . 77±1 . 99 µm , n = 7 ) . This analysis suggests that the parallel bundles of microtubules seen spanning the entire width of the epithelium within individual neuroepithelial cells play a critical role in the ability of cells to maintain their length and proper tissue architecture , as they do in homeostatic length control in cells in culture .
Taken together our analysis of hemocyte-derived Drosophila cells and epithelial-derived HeLa cells on micro-contact-printed lines in culture , and of neuroepithelial cells in the developing zebrafish neural tube , identifies a capacity for animal cells to maintain an intrinsically defined length . Cell length control in these systems appears to act independently of cell width and volume . Instead , it relies on the continuing presence of a polarised population of dynamic microtubules that , as a result of interactions with the cell cortex , come to lie parallel to the long axis of cells , running from the centre to either cell tip . Because cell extension depends on microtubules , this oriented array of dynamic microtubules is then in a position to regulate cell length . Somewhat surprisingly , we observed no role for actin-based cortical tension in opposing microtubule-driven cell elongation in our system . This is in line with data from previous studies showing that the final cell spread area is independent of the actin cytoskeleton [10] , [24] . Because of this , cell length control is unlikely to reflect a balance of forces between contractile actin filaments and extending microtubule rods—as previously hypothesized for the regulation of animal cell form [22] . In fact , a recent study showed that microtubules do not bear a significant mechanical load in fully spread cells [53] . Based on these observations , it seems likely that microtubules drive cell elongation through the promotion of directional traffic of material from the Golgi to the cell surface [24] , rather than through the direct generation of mechanical force itself . Interestingly , microtubule dynamics have also been shown to contribute to ( i ) the regulation of cell length and cell shape in fission yeast via the regulation of cell transport [54] , ( ii ) axial elongation in plant cells [55] , and ( iii ) the maintenance of spindle length within defined limits [56] . As such , limits to the length of dynamic microtubule-based structures may be a relatively widespread phenomenon in biology . Recently , a mechanism for length-dependent microtubule depolymerisation was discovered [5] , which could help to explain the regulation of processes such as cell length . While this mechanism may be involved in cell length control , our analysis suggests that it may not be necessary to invoke this type of control to understand the maintenance of microtubule-dependent structures within relatively well-defined size limits in all cases , since cell length control can arise as a relatively simple by-product of microtubule-based cell extension if a few simple propositions hold . These propositions are the following: ( i ) that dynamic microtubules are polarised so that they polymerise along the long cell axis towards cell tips , ( ii ) that the rate of microtubule catastrophe increases when microtubules reach the cell's ends , and ( iii ) that microtubules act together to prevent retraction of the cell margin and promote cell elongation , probably through the delivery of material to counter turnover of material at the cell tips . Significantly , a model based on these experimentally well-established assumptions predicts the spreading dynamics we observe on patterned lines , the steady-state microtubule distribution ( Figure 7D ) , the path of cell elongation ( Figures 9A , 9B , and S4B ) , and the effects of a microtubule poison on cell length ( Figure 9C and 9D ) . Intrinsic regulation of cell length in a tissue and developmental context is inherently difficult to demonstrate unambiguously because of the potentially confounding effects of forces and signals from neighbouring tissues and extracellular matrix . For example , tissue architecture and the dimensions of neuroepithelial cells in the newly formed zebrafish neural tube are likely to be influenced by both extrinsic and cell-intrinsic factors , as cells undergo interdigitation and intercalation across the midline [16] , [57] . Moreover , zebrafish neuroepithelial cells achieve their final length through a complex mechanism involving initial overextension and then retraction ( data not shown ) as they establish a nascent apical domain at the tissue midline [16] . Because tissue architecture in this case is generated via a complex process of self organisation , an intrinsic mechanism that biases cell length may be indispensable to ensure robust organ size and form in the face of variations in the volume of component cells and variations in the movements and differentiation of surrounding tissues [58] . Indeed , here we show that neuroepithelial cell length in vivo is independent of cell volume but , consistent with our findings in vitro , dependent on a population of axially oriented microtubules . As a result , increasing levels of microtubule inhibitors administered over a relatively short period of time cause systematic reductions in epithelial height ( Figure 10C ) . These data help to explain how it is that neuroepithelial height can remain relatively unchanged in ectopic neural tubes that are not situated at the embryonic midline in morphogenetic mutants ( compare cells in Figure 4C and 4F in [16]; compare cells in Figure 1G , 1H , and , 1K in [57] ) . Similar conclusions can be drawn from earlier studies on the effects of cell cycle arrest on development [59] , [60] , where it was noted that many ultrastructural features and functional properties of cells were conserved despite dramatic changes induced in cell size . Our study shows that the regulation of neuroepithelial height is one way by which embryos are able to do this . More generally , one could hypothesize that cell length homeostasis is likely to be required in all growing epithelia that need to maintain apical–basal structure despite continual changes in the volumes of their constituent cells . While the emphasis of this study is cell length homeostasis , it is clear that changes in animal cell length or height are likely to play a critical role during tissue morphogenesis in vivo . During Drosophila gastrulation , for example , epithelial cells are thought to actively contribute to ventral furrow formation by undergoing apical constriction and cell shortening [15] , a process that would seem to require orchestrated changes in cell length . Although the mechanism by which this occurs is not understood , it is plausible that microtubules , which are highly polarised along the apical–basal cell axis of Drosophila epithelial cells , play a role , as has been shown in other epithelia [61] , [62] . Similarly , the changes in cell length that accompany neuron and myoblast differentiation are brought about by changes in microtubule dynamics [63]–[65] . Thus , many animal cells are likely to need to be able to regulate their optimal steady-state length , e . g . , by changing the rate of microtubule polymerisation to alter their resting length whilst preserving length control ( Figures 9A and S4A ) . Conversely , decreasing the susceptibility of microtubules to undergo catastrophe at the cell cortex , e . g . , by crosslinking microtubules with Map1A [63] , could induce dramatic but relatively unregulated cell elongation ( Figure 8E ) . Because of this , long cells like neurons may rely on additional control systems , like length-dependent regulators of microtubule catastrophe [5] or environmental cues , to reach specific locations during the process of axon pathfinding . An important goal of future research will be to identify how intrinsic length constraints and additional layers of control are altered during animal development to give specific cells and tissues their characteristic forms , and to determine whether the deregulation of cell length control contributes to the loss of tissue homeostasis seen in diseases such as cancer .
A mathematical model was formulated to examine the likely role of microtubules in the control of cell length . The model was based on experimental observations that microtubules aligned along the long axis of the cell , with plus ends towards the cell tips ( Videos S1 , S2 , S3 ) , appear to drive the elongation of cell edges on adhesive lines ( Figure 4A–4F ) and in cells lacking lamellipodia ( Figure 3 ) . Using this as a framework , we constructed a simple stochastic half-cell model of cell elongation driven by a population of parallel dynamic microtubules . Where possible , parameters used to model microtubule dynamics were taken directly from experimental data ( see Table 1 ) . We then made a number of simplifying assumptions . First , we assumed that a fixed number of microtubules , Nm , nucleate at the cell centre and grow towards the cell ends at a rate determined by the known rate of microtubule polymerisation , vg , with an experimentally defined cytoplasmic catastrophe rate cI [43] , and we made the simplifying assumption that there is no rescue of microtubule growth following catastrophe . Upon reaching the cortex ( defined as the region within 3 µm of the cell boundary ) , microtubule plus ends then act together to promote extension of the cell boundary . We modelled the cooperative effects of n microtubules touching the cortex in driving cell elongation at each time step using the function e−α/n . Because microtubules can drive cell elongation through forces generated by the addition of tubulin subunits [70] , through the delivery of new material required for local growth , and/or through local modification of the cell cortex [45] , [46] , in this study α is assumed to be a free , dimensionless parameter . At the same time , in line with cell biological data , contact with the cortex induces an increase in the rate of microtubule catastrophe cB [43] ( See Table 1 ) . Since catastrophe events free up a microtubule nucleation site in the model , the number of growing microtubules remains constant over time . Finally , the term vB was added , based upon experimental data ( data not shown ) , to represent the slow retraction of the cell margin in the absence of microtubules ( 0 . 4±0 . 3 µm/min ) . After applying known rates of microtubule growth and catastrophe to the model ( see Table 1 ) , two free variables remain: α , which governs the cooperative effect of microtubules on the movement of the cell boundary , and Nm , the number of microtubule nucleation sites . In order to compare the results of simulations with experimental data from Figures 1–3 , simulations were run in Matlab by using parameter values shown in Table 1 , and the position of the boundary was used as a read out of half cell length . Based on this scheme , given a random number r in the interval [0 , 1] , the stochastic equations that describe the microtubule growth in time are given by ( 1 ) Where Lm is the microtubule length , r is a random number in the interval 0≤r≤1 , Δt ( 0 . 001 min ) is the simulation time step , and vm and cm are the microtubule velocity and catastrophe rates , respectively . The stochastic catastrophe event is defined by the random number r and Δt cm . In order to model cell extension , a cell length boundary equation LB is added to the model: ( 2 ) where vg is the microtubule growing rate internal to the cell and vB is the cell boundary velocity retraction rate when there are no microtubules crossing the boundary . α governs the cooperative effect of microtubules in promoting cell boundary extension , as defined by the function e−α/|mB ( t ) | . mI and mB describe internal and boundary microtubules as follows: ( 3 ) ( 4 ) Finally the general microtubule velocity and catastrophe rate equations are given by ( 5 ) ( 6 ) where cI and cB are the internal and boundary catastrophe microtubules rates , respectively . The aim of our model was to quantitatively explore the relationship between microtubule behaviour and cell length control . Significantly , for a wide range of Nm and α values , this simple scheme recapitulated the path of cell elongation and length control seen in observations of cells on lines ( Figure 1 ) . As seen in experiments ( Figure 1D ) , the rate of cell elongation in the model diminished over the course of 60 min , leading to a steady-state cell length within a few hours ( Figure 7B ) . To understand the source of cell length homeostasis in these simulations we plotted the number of microtubules contacting the cell cortex over time ( Figure 7C ) . This revealed a steady decrease in the number of microtubules reaching the cell cortex as cells elongate . As cells extend , this number plateaus , reaching a steady equilibrium between cell elongation and cell retraction that maintains cell length over time , which is typically approximately two microtubules—a number that is independent of the number of microtubule nucleation sites and cell length itself ( Figure 7C ) . The model also predicted a linear decrease in microtubule density with distance from the cell centre similar to that measured in cells on lines ( Figure 7D ) . Significantly , the number of microtubule nucleation sites , Nm , had little impact on the ability of cells to achieve length homeostasis ( Figure 8A and 8B; cell length variance is used as quantitative measure of homeostasis ) , while a moderate level of microtubule cooperation ( α>3 ) was required for a reproducible cell length ( Figure 8B and 8C ) . Above this threshold , while cells were able to maintain a homeostatic cell length irrespective of the specific values of Nm and α , cell length increased with increasing values of Nm and decreased with increasing values of α ( Figure 8C ) . Although catastrophe rates used in the model were based on experimentally well-defined parameters , we also determined the effects of changing the cortical cB and internal cI catastrophe rates on cell length homeostasis . This revealed that cell length control is gradually lost when cB/cI tends to zero , i . e . , as cB values were reduced or cI increased . This is seen by the increase in cell length variance—a quantitative measure of homeostasis ( Figures 8D and S4B ) . It should also be noted that at values of cB close to zero , the number of microtubules at the cell boundary increases to a high steady-state value , driving continuous cell elongation ( Figure 8E ) . Finally , the model was used to test the likely effects of colcemid in this system [38] by altering vg . A reduction in vg leads to a linear reduction in cell length ( Figure S4A ) , as cells re-establish equilibrium with an average of approximately two microtubules contacting the cell cortex per unit time ( Figure 9A and 9B ) .
|
Because many physical processes change with scale , size control is a fundamental problem for living systems . While in some instances the size of a structure is directly determined by the dimensions of its individual constituents , many biological structures are dynamic , self-organising assemblies of relatively small component parts . How such assemblies are maintained within defined size limits remains poorly understood . Here , by confining cells to spread on lines , we show that animal cells reach a defined length that is independent of their volume and width . In searching for a “ruler” that might determine this axial limit to cell spreading , we identified a population of dynamic microtubule polymers that become oriented along the long axis of cells . This growing population of oriented microtubules drives extension of the spreading cell margin while , conversely , interactions with the cell margin promote microtubule depolymerisation , leading to cell shortening . Using a mathematical model we show that this coupling of dynamic microtubule polymerisation and depolymerisation with directed cell elongation is sufficient to explain the limit to cell spreading and cell length homeostasis . Because microtubules appear to regulate cell length in a similar way in the developing zebrafish neural tube , we suggest that this microtubule-dependent mechanism is likely to be of widespread importance for the regulation of cell and tissue geometry .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"cell",
"biology/cell",
"growth",
"and",
"division",
"cell",
"biology",
"computer",
"science/systems",
"and",
"control",
"theory",
"cell",
"biology/cell",
"adhesion",
"cell",
"biology/cytoskeleton"
] |
2010
|
A Polarised Population of Dynamic Microtubules Mediates Homeostatic Length Control in Animal Cells
|
Chemotherapy is still a critical issue in the management of leishmaniasis . Until recently , pentavalent antimonials , amphotericin B or pentamidine compounded the classical arsenal of treatment . All these drugs are toxic and have to be administered by the parenteral route . Tamoxifen has been used as an antiestrogen in the treatment and prevention of breast cancer for many years . Its safety and pharmacological profiles are well established in humans . We have shown that tamoxifen is active as an antileishmanial compound in vitro , and in this paper we analyzed the efficacy of tamoxifen for the treatment of mice infected with Leishmania amazonensis , an etiological agent of localized cutaneous leishmaniasis and the main cause of diffuse cutaneous leishmaniasis in South America . BALB/c mice were infected with L . amazonensis promastigotes . Five weeks post-infection , treatment with 15 daily intraperitoneal injections of 20 mg/kg tamoxifen was administered . Lesion and ulcer sizes were recorded and parasite burden quantified by limiting dilution . A significant decrease in lesion size and ulcer development was noted in mice treated with tamoxifen as compared to control untreated animals . Parasite burden in the inoculation site at the end of treatment was reduced from 108 . 5±0 . 7 in control untreated animals to 105 . 0±0 . 0 in tamoxifen-treated mice . Parasite load was also reduced in the draining lymph nodes . The reduction in parasite number was sustained: 6 weeks after the end of treatment , 1015 . 5±0 . 5 parasites were quantified from untreated animals , as opposed to 105 . 1±0 . 1 parasites detected in treated mice . Treatment of BALB/c mice infected with L . amazonensis for 15 days with tamoxifen resulted in significant decrease in lesion size and parasite burden . BALB/c mice infected with L . amazonensis represents a model of extreme susceptibility , and the striking and sustained reduction in the number of parasites in treated animals supports the proposal of further testing of this drug in other models of leishmaniasis .
Protozoan parasites of Leishmania genus are the etiological agents of leishmaniasis , a disease distributed worldwide with a broad spectrum of clinical manifestations according to the causative species and immunological status of the host . Leishmaniasis current therapy is mainly based on the systemic administration of toxic pentavalent antimonials or amphotericin B , drugs with several side effects , such as arrhythmia , nephro- and hepatotoxicity . Additionally , emergence of Leishmania strains resistant to antimonials has been reported [1] , [2] . Recently , miltefosine has been approved in India for the therapy of visceral leishmaniasis [3] , but its efficacy on the treatment of American cutaneous leishmaniasis has been shown to be variable depending on the causative species [4] , [5] , [6] , [7] . Therefore , new alternatives for the treatment of leishmaniasis are greatly needed . In South America , Leishmania amazonensis is one of the causative agents of localized cutaneous leishmaniasis and the most important agent of diffuse cutaneous leishmaniasis ( DCL ) , a devastating disease with uncontrolled progression , characterized by multiple skin lesions and vaste numbers of amastigotes . As a rule , there is no satisfactory response to DCL treatment [8] , [9] . We have previously shown that the antiestrogen tamoxifen , a drug extensively used as a chemotherapeutic and chemopreventive agent against breast cancer , presents leishmanicidal activity in vitro . This drug has a direct leishmanicidal effect and it also shifts the pH of parasitophorous vacuoles from acid to neutral , which in turn heightens the drug effect on amastigotes . Tamoxifen concentrations of approximately 10 µM inhibit 50% of L . amazonensis viability and growth in vitro [10] . In the present study we demonstrate that L . amazonensis-infected BALB/c mice treated with tamoxifen for 2 weeks presented a significant reduction in lesion size and parasite burden .
The Ethics Committee that has approved this study is the Ethics Committee for Animal Experimentation of the Instituto de Ciências Biomédicas , University of São Paulo . L . amazonensis promastigotes ( MHOM/BR/1973/M2269 ) were grown in Medium 199 ( Sigma-Aldrich ) supplemented with 10% heat-inactivated fetal calf serum ( FCS; Invitrogen ) and incubated at 25°C . Female BALB/c mice ( 4–5 week-old ) were inoculated with 5×106 stationary-phase parasites at the base of the tail . Five weeks after infection , mice were randomly assigned into experimental groups ( n = 7–10 ) . Treated groups received intraperitoneal injections of 30 . 4 mg tamoxifen citrate/kg/day ( the drug equivalent to 20 mg/kg/day tamoxifen ) or 20 mg/kg/day meglumine antimoniate ( Glucantime ) for 15 days . Tamoxifen citrate was purchased from Sigma-Aldrich , USA; Glucantime was a kind gift from Sanofi-Aventis . Stock solutions of tamoxifen were prepared in saline every two days and stored at 4°C . Disease progression was evaluated once a week by recording the average diameter of the tail measured as the mean of tail base diameters in horizontal and vertical directions ) and the ulcer size , expressed as the ulcer area in mm2 . Measurements were taken with a caliper ( Mitutoyo Corp . , Japan ) . Body and uterus weights were also registered . Animal experiments were repeated four times and were approved by the Ethical Committee . Parasite burden from infected tissue was quantified as described previously [11] . Promastigotes differentiated from lesion amastigotes were used on drug sensitivity assays while in the first passage in vitro . Cellular viability was assessed by measuring the cleavage of 3- ( 4 , 5-dimethylthiazol-2-yl ) -2 , 5-diphenyl tetrazolium bromide ( MTT; Sigma-Aldrich ) by metabolically active cells as described [12] . Data on lesion progression were analyzed for statistical significance by using the non-parametric Mann-Whitney test ( GraphPad Prism 5 software ) . Results of limiting dilution assay were analyzed based on two-tailed Student t test for paired samples using the ELIDA software . A result was considered significant at P<0 . 05 .
The treatment of L . amazonensis-infected BALB/c mice was initiated 5 weeks post-infection , time when lesions were already established and apparent . Mice were treated with 20 mg/kg/day tamoxifen intraperitoneally for 15 days . No toxic effects were detected during or after drug treatment . At the end of treatment , the average body weight in animals treated with tamoxifen was equivalent to values for the control group ( untreated mice: 26 . 2±1 . 1 g; treated mice: 25 . 4±1 . 8 g ) and the average weight of uteri indicated no significant alteration between tamoxifen-treated ( 0 . 18±0 . 4 g ) and untreated mice ( 0 . 21±0 . 6 g ) . Figure 1A shows the progression of lesion size in untreated versus tamoxifen-treated mice . During and after tamoxifen administration we observed that treated animals presented less swelling at the infection site when compared to control animals . A statistically significant difference between the average thickness of lesions of untreated and tamoxifen-treated mice was evident on completion of treatment , at week 7 post-infection ( P<0 . 001 ) and remained clear until the end of the experiment ( week 13 , P<0 . 01 ) , when control mice had to be euthanized . Macroscopical aspects of the lesions are displayed in Figure 1B for untreated ( left column ) and tamoxifen-treated animals ( right column ) . Since L . amazonensis infection in BALB/c mice normally evolves from swelling at the infection site to an ulcerated lesion with loss of tissue , the measurement of lesion thickness can be misleading at late time-points . So , another criteria used for evaluating disease progression was the appearance and enlargement of ulcers . Tamoxifen treated mice showed a very significant delay in the development of ulcers when compared with untreated mice ( Figure 1C ) . Parasite burden in tamoxifen-treated animals was evaluated immediately after the interruption of treatment ( 7 weeks after infection ) and 6 weeks later ( 13 weeks after infection ) . As shown in Figure 2 , a significant decrease on total parasite numbers per lesion on tamoxifen-treated animals was observed in both time points . At the end of the experiment , the average number of parasites was reduced by at least 99 . 7% in treated groups , as compared to untreated animals . These results were reproduced in 3 independent experiments . In order to evaluate the activity of tamoxifen in parallel with a standard drug , a fourth experiment was performed with groups of 10 mice treated with 20 mg/kg/day tamoxifen , 20 mg/kg/day Glucantime or mock-treated with saline . Treatment was initiated 4 weeks after infection and carried on for 15 days with daily intraperitoneal infections . As shown in Table 1 , three weeks after the end of treatment , there was no difference in the average size of lesions between mock and Glucantime-treated mice . The group that received tamoxifen showed a significant decrease in lesion thickness . Parasite burden was determined for the lesion site , draining lymph node and spleen . There was a significant reduction in the numbers of parasites recovered from tamoxifen treated mice as compared to mock or Glucantime groups both at the lesion and lymph node . No parasites were recovered from the spleen in any of the groups ( Table 1 ) . Therefore , tamoxifen proved to be more effective in this experimental model than the standard drug . Finally , we investigated whether parasites remaining in tamoxifen treated groups were less sensitive to the drug . MTT viability assays showed that tamoxifen's activity against promastigotes derived from parasites extracted from treated or untreated mice remained unchanged with EC 50% of 11 . 5±1 . 1 and 12 . 8±2 . 8 µM , respectively . Therefore , remaining parasites did not develop resistance to tamoxifen during treatment .
Our data reveal a significant effect of tamoxifen in the reduction of skin lesions caused by L . amazonensis in BALB/c mice . Effectiveness was apparent not only as reduced swelling and ulceration in treated animals but also as an important reduction in parasite burden . The experimental model of infection used in this study is one of extreme susceptibility . BALB/c mice infected subcutaneously with L . amazonensis develop progressive swelling at the inoculation site , followed by ulceration and loss of tissue simultaneous with the appearance of methastasis at distant sites . The treatment did not lead to sterile cure of lesions but Leishmania parasites have been shown to remain present and viable , although in decreased numbers , after treatment with antimonials in a variety of animal models , as well as in humans . The lack of clinical or parasitological response to Glucantime in L . amazonensis BALB/c infected mice has been reported previously [13] . Furthermore , the timing for initiation of treatment can significantly influence the disease outcome , as stressed by previous studies [14] . Our experimental treatments were initiated 30–35 days after mice infection , an interval of time that allowed the establishment of disease and when infection sites were already swollen and , in some animals , had started ulcerating . So , the data shown here imply that the intraperitoneal administration of tamoxifen resulted in a remarkable response to treatment . We are currently evaluating tamoxifen's efficacy in the treatment of other models of cutaneous and visceral leishmaniasis . We have shown in vitro that tamoxifen leishmanicidal effect is independent of the estrogen receptor [10] and therefore it is unlikely that response to treatment would be different in male or female mice . Indeed , preliminary results obtained in Leishmania chagasi infected hamsters show no gender-related effect on the anti-leishmanial response to tamoxifen . Apart from its direct leishmanicial activity , tamoxifen mode of action in vivo could involve other pathways favouring amelioration of the infection . Tamoxifen has been reported to increase synthesis of inducible nitric oxide synthase and production of nitric oxide [15] . We did not detect differences in the accumulation of nitrate on supernatants of L . amazonensis infected macrophages treated or not treated with tamoxifen ( data not shown ) . The metabolite profile of tamoxifen varies in different animal models [16] . This drug has been used in mice in a variety of doses and administration schemes . The dosage employed in this study was chosen based on previous reports showing that , in mice , daily intraperitoneal injections of 25 to 100 mg/kg of tamoxifen resulted in drug serum levels similar to those observed in patients [17] . This antiestrogen has been widely used for treatment and prevention of breast cancer [18] . The most serious side effect observed on clinical grounds is an increased risk for endometrial cancer which appears after prolonged use . Effects observed in our experiments suggest that antileishmanial therapy with tamoxifen would not require extensive periods of treatment . We did not detect changes in uterine weight in treated female mice , a well-established parameter for evaluation of tamoxifen's toxicity [19] . Tamoxifen administered at 40 mg/kg/day for 4 weeks has been recently shown to impair bone growth in rats [20] raising concerns on the application of this drug to treat leishmaniasis in children . We are also investigating the effect of other selective estrogen receptor modulators with different effects in bone metabolism , like raloxifene , as antileishmanial drugs . The potential value of tamoxifen for treating human leishmaniasis needs further evaluation . To the best of our knowledge , this is the first report of an in vivo investigation on tamoxifen's efficacy against Leishmania infection and points to a new alternative in the treatment of leishmaniasis .
|
Leishmaniasis is an antropozoonotic disease with a wide range of clinical manifestations . In humans , signs of disease vary from skin and mucosal ulcers to enlargement of internal organs such as the liver and spleen . The unicellular parasite Leishmania amazonensis is able to infect humans and cause localized or diffuse skin lesions . The treatment for this disease is difficult , as it requires prolonged and painful applications of toxic drugs that are poorly tolerated . Therefore , a key area in leishmaniasis research is the study of new therapeutic schemes and less toxic drugs . The present report is based on the investigation of tamoxifen's activity ( a compound that has been in clinical use since the 1970s for the treatment of breast cancer ) in the treatment of mice experimentally infected with L . amazonensis . We observed that infected mice treated with 20 mg/kg/day of tamoxifen for 15 days showed a significant clinical and parasitological response , with reduction in the size of lesions and ulcers and decreased numbers of parasites . These promising results pave the way for further testing of this drug as a new alternative in the chemotherapy of leishmaniasis .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"infectious",
"diseases/protozoal",
"infections",
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"infectious",
"diseases/antimicrobials",
"and",
"drug",
"resistance"
] |
2008
|
Tamoxifen Is Effective in the Treatment of Leishmania amazonensis Infections in Mice
|
Cholesterol Dependent Cytolysins ( CDCs ) are important bacterial virulence factors that form large ( 200–300 Å ) membrane embedded pores in target cells . Currently , insights from X-ray crystallography , biophysical and single particle cryo-Electron Microscopy ( cryo-EM ) experiments suggest that soluble monomers first interact with the membrane surface via a C-terminal Immunoglobulin-like domain ( Ig; Domain 4 ) . Membrane bound oligomers then assemble into a prepore oligomeric form , following which the prepore assembly collapses towards the membrane surface , with concomitant release and insertion of the membrane spanning subunits . During this rearrangement it is proposed that Domain 2 , a region comprising three β-strands that links the pore forming region ( Domains 1 and 3 ) and the Ig domain , must undergo a significant yet currently undetermined , conformational change . Here we address this problem through a systematic molecular modeling and structural bioinformatics approach . Our work shows that simple rigid body rotations may account for the observed collapse of the prepore towards the membrane surface . Support for this idea comes from analysis of published cryo-EM maps of the pneumolysin pore , available crystal structures and molecular dynamics simulations . The latter data in particular reveal that Domains 1 , 2 and 4 are able to undergo significant rotational movements with respect to each other . Together , our data provide new and testable insights into the mechanism of pore formation by CDCs .
Cholesterol dependent cytolysins ( CDCs ) represent a major branch of the CDC/membrane attack complex/perforin-like ( MACPF ) protein superfamily . Originally identified as virulence factors produced by Gram positive pathogens , CDC toxins have recently been identified in Gram negative bacteria such as Desulfobulbus propionicus and Enterobacter lignolyticus [1] , [2] . Well-studied family members include perfrinolysin O ( PFO ) , pneumolysin ( PLY ) , listerolysin O ( LLO ) , streptolysin O ( SLO ) and intermedilysin ( ILY ) . A unifying feature of these toxins is the ability to assemble into giant , membrane embedded pores [1] . Pore formation is associated with a variety of toxic functions , including escape from the intracellular phagolysosome ( LLO ) [3] and the delivery of folded toxins such as nicotinamide adeninedinucleotide-glycohydrolase by SLO [4] . The structure of CDCs has been well studied . The first crystal structure of a monomeric CDC ( PFO ) suggested that the molecule comprises four distinct domains . Domains 1 and 3 are non-contiguous regions forming a head region that is linked via Domain 2 to the Ig-like Domain 4 ( Figure 1A ) [5] . The mechanism of CDC membrane insertion has also been well characterized and mapped to the structure . Briefly , during pore formation two clusters of helices ( Transmembrane Helix 1 ( TMH1 ) and 2 ( TMH2 ) ) within Domain 3 unwind and insert into the membrane as two amphipathic β-hairpins . Together , Domains 1 and 3 are homologous to the distantly related MACPF proteins [6] . Domain 2 , a region unique to CDCs , essentially comprises an elongated three-stranded β-sheet that links the pore forming head domain ( Domains 1 and 3 ) to Domain 4 . Finally , Domain 4 contains the determinants for interacting with the membrane , including a key conserved sequence that is important for binding cholesterol [7] . Single particle cryo-Electron Microscopy ( SP-cryo-EM ) , Atomic Force Microscopy and Förster Resonance Energy Transfer ( FRET ) studies [8]–[10] have provided key insights into the transition from the prepore to the pore structure . Following interaction with the membrane surface via Domain 4 , CDC monomers assemble into a prepore form . In this conformation , SP cryo-EM data suggest that the conformation of each subunit broadly resembles that seen in crystal structures ( i . e . no major conformational change is apparent ) . Biophysical and microscopy data reveal that following prepore assembly , and in order to form a transmembrane pore , Domains 1 and 3 undergo a significant 40 Å movement towards the membrane surface [9] , [10] . Further , the cryo-EM structure of the pneumolysin pore [8] shows that the central four-stranded β-sheet opens , an event that separates Domains 2 and 3 . Concomitant with these events , the two small clusters of α-helices TMH1 and TMH2 on either side of the central sheet unwind and insert into the membrane as amphipathic β-strands ( Figure 1B ) . The conformational changes that surround Domains 1 and 3 are relatively well understood . However , a key question remains about how the prepore form collapses towards the membrane surface . Interpretation of cryo-EM data strongly suggests that Domain 2 “buckles” or “doubles over” itself . However , these data are of low resolution ( 29 Å ) and to date it has not been possible to unambiguously model the position and conformation of Domain 2 [8] . Furthermore , attempts at conformationally trapping Domain 2 to prevent buckling have been unsuccessful [11] . Therefore understanding the structural perturbations that take place in Domain 2 remains central to understanding the mechanism of membrane insertion in CDCs . Previous crystallographic studies have demonstrated wide variability in the position of the membrane binding Domain 4 with respect to Domains 1 , 2 and 3 [12] , [13] . It has also been suggested that Domain 2 distortion governs different orientations of Domain 4 [12] . In contrast , a second hypothesis postulates that movement in Domain 4 is entirely attributable to a hinge bending motion located at the Domain 2/4 interface [11] , [14] . However , to date , there has been no family-wide description of the regions of rigidity and plasticity of the CDCs . Here , we characterize the variability between the fifteen available CDC crystal structures and use this information to re-visit the role of Domain 2 in conformational change using the published cryo-EM maps [8] This analysis allowed a novel and methodical molecular model building strategy . Our data suggest that a rotational collapse involving Domain 2 provides the most logical mechanistic model for CDC pore formation with the current available data .
To characterise the rigid fragments we performed superposition experiments [15] on all known CDC crystal structures ( Table 1 ) . By first aligning the whole molecules we identified a major rigid body consisting of Domains 1 , 3 ( excluding the TMH2 region ) and the upper part of Domain 2 close to Domain 1 and packing against TMH1 ( positions 53–56; 81–90; 380–384 ) ( Figure 2A , B ) . The alignment highlights the structural variability of the base of Domain 2 as well as the variability of Domain 4 orientation across the family ( Figure S1 ) . Further , structural alignments of Domains 2 and 4 either separately or together ( Figure 2C , D ) demonstrate that Domain 4 is to be treated as a rigid body and identify the base of Domain 2 as a region of plastic deformation . Closer inspection of the structural alignments identified a direct relationship between the deformation of residues in the base of Domain 2 ( positions 69–76; 387–390 ) and the orientation of Domain 4 . As the overall twist in Domain 2 β-sheet increases Domain 4 rotates away from the body of the molecule by up to 35° ( Figure 3 ) . This defines that the orientation of Domain 4 in CDCs is in part attributable to the plasticity at the base of Domain 2 . We next investigated the deformation in PFO Domain 2 . This is made possible by the availability of multiple structures that allow us to define the conformational space accessible to one molecule . We identified seven conformations derived from three crystal forms ( Table 1 and [16] ) . In two conformations ( PFO IIIA and C ) the interface between Domains 2 and 3 is partially disrupted . PFO IIIA and IIIC conformers display a loss of contacts between the TMH2 region and Domain 2 , corresponding to a loss of ∼180 Å2 ( 30% ) surface area . This loss of surface area is associated with an increased distance between the pair of residues Ser287 ( TMH2 ) and Glu388 ( Domain 2 ) compared to the other five conformers ( Table 2 and [16] ) . Notably these contacts all involve amino acids located at the base of Domain 2 ( Ile76 , Ser386 & Glu388 ) that we have identified as a region of plasticity ( Figure 4D ) . Although it has been identified that the loss of contacts at the Domains 2/4 interface is associated with distortion of the elongated Domain 2 β-sheet [16] , little characterization of this region has been performed . Consequently we focused on the deformation in PFO Domain 2 by measuring the twist of the β-sheet in terms of inter-strand pairing of amino acids upon partial loss of this interface ( Figure 4C , D ) . The twist values represented in Figure 4C show moderate differences as they mostly overlap and are characterized by two peaks at positions 382 ( high twist central to the sheet ) and 390 ( C-terminal region of the sheet ) ( Figure 4D ) . However , we note a reduced twist in conformations with a weaker interface ( orange ) with differences of 6° and 11° at two consecutive positions: Thr384 and Thr385 . A lower twist of the sheet at these positions immediately comprised between the major central twist and residues involved in the TMH2-Domain 2 interface leads to a significant shift of ∼8 Å of the C-terminal segment of Domain 2 ( Figure 4D ) . This is indicative that the partial loss of the interface is structurally coupled with un-twisting of Domain 2 . As crystallographic packing artifacts may limit our structural analysis we performed a series of molecular dynamics ( MD ) simulations starting from conformations with either a full ( PFO I×2 , PFO IIIB×1 , summarized in Table 2 ) or partial ( PFO IIIA×1; PFO IIIC×1 ) Domains 2/TMH2 interface . We observed that the molecule has the ability to transition between full and partial interface ( PFO I Simulation 2 , PFO IIIB , Figure 5 and Videos S1 , S2 , S3 , S4 , S5 ) . Conversely PFO IIIC was able to transition from a partial interface to a full interface . Associated with the fluctuations of the interface , we observed a clustering of twist values at the positions outlined by our structural analysis ( Thr384-Arg80 and Thr385-Glu79 , Figure 4C , Table 2 ) . An increase in the distance between TMH2 and Domain 2 ( i . e . from full to partial interface ) is associated with a decrease in the twist values ( Figure 5 ) . We therefore conclude from MD simulations that PFO has the potential to fluctuate between discrete states independent of crystal packing . In addition , structural analysis of the MD simulations identified a common pattern whereby the partial loss of the interface is structurally coupled with un-twisting , or straightening , of Domain 2 . Next we analyzed our MD simulations in terms of the positions adopted by Domain 4 . In all simulations performed , Domain 4 exhibited a large range of orientations ( Figure 6A , Figure S2 ) . Principal component analysis of the MD simulations identified such domain movement as the slowest ‘breathing’ mode of motion with a minimum of the total variance explained of 40% and a collectivity of typically 0 . 6 across all simulations . Furthermore , in simulations where the Domain2/TMH2 interface transitions between full and partial states ( PFO I simulation 2 , PFO IIIB ) we found that a single mode ( the third slowest mode in both cases ) best described the departure of the base of Domain 2 from TMH2 . Notably , this mode also encapsulated accompanying motion of Domain 4 , rotating away from the body of the molecule ( Figure 6B ) . Thus , MD simulations further support that the orientation of Domain 4 in PFO can be ascribed to the plasticity at the base of Domain 2 . To extend our investigation of the conformational properties of CDCs we performed additional MD simulations of other members of the family including intermedilysin , suilysin , anthrolysin and streptolysin . The major observation is the consistent flexibility of Domain 4 with respect to domains 1–3 , similar to the flexibility observed for PFO ( Figure 6A , Figure S2B , Videos S6 , S7 ) . Secondly , for ILY IA and to a smaller extent for SLY simulations we observed a reduction of twist where the Domain 2/TMH2 interface is partially lost ( Figure S2A ) . This correlates with the observations from the PFO simulations . We also noted that partial loss of the interface was associated with a similar decrease in twist at nearby positions ( ILY IA ( 105–412 ) and SLY ( 73–383 ) , not shown ) . This data suggests the torsion of the β-sheet may be subtly modulated in a toxin-dependent fashion . Overall these data , although non-exhaustive , support that two characteristics emerge as unifying features across the CDC family: the substantial relative flexibility of Domain 4 and the ability of Domain 2 to straighten upon weakening of the Domain 2/TMH2 interface . Taken together , our cross-comparison of all available CDC structures and MD simulations analysis allow us to define the conformational properties of the molecule in the monomeric form . Domain 2 wraps around TMH1 and the base of TMH2 and most likely prevents their premature release and aggregation of the molecule ( Figure 1A and as demonstrated in the case of LLO [17] ) . Our data suggest that preservation of the interface of TMH1 and TMH2 with Domain 2 is accompanied by conformational torsion in Domain 2 and twisting of the elongated sheet . We suggest that release of this interface , an early and critical step of membrane insertion [18] , results in the straightening of Domain 2 . Domain 2 has been first proposed to undergo some conformational change after loss of TMH1/2 contacts in order to account for the 40 Å vertical collapse observed upon prepore-to-pore transition [9] . Tilley and coworkers [8] hypothesized that one way to account for this collapse was for the triple stranded β-sheet of Domain 2 to fold sharply in half . Here , and in contrast to this idea , our combined structural analysis and MD simulations suggest that Domain 2 has the propensity to straighten upon the loss of TMH1/2 contacts . Moreover , we argue that an energetic requirement to preserve inter-strand hydrogen bonds and local packing in anti-parallel β-sheets favours continuous deformation of the Domain 2 region rather than a major collapse [19] . Given this analysis , we re-visited the conformational states of prepore and pore in CDCs with improved knowledge of their conformational properties . Next we examined the conformation of the PLY monomer within the prepore oligomer . We modeled the prepore conformation within the available cryo-EM map with 31-fold circular symmetry ( C31 ) [8] ( Figure 7A , B; detailed in Methods ) . After the flexible fitting step the cross-correlation coefficient ( CCC ) for the oligomer improved from 0 . 57 to 0 . 61 . The structural transitions accompanying the assembly of the prepore are well described . First the conformationally labile β5 strand rotates away from the β4 strand leaving its edge exposed to the formation of mainchain hydrogen bonds with the β1 strand of an adjacent monomer [20] . Our model takes this structural change into account . The β5 strand is modeled here as a short helix by analogy with the structurally equivalent position in the complement component C6 [21] ( Figure S3 ) , a member of the distantly but structurally and functionally related MACPF family [6] , [22] . Secondly , the oligomer transitions to a SDS-resistant prepore upon the formation of specific β1–β4 contacts [20] , [23] . Our prepore model displays some , but not all , oligomeric β1–β4 mainchain hydrogen bonds compatible with the pattern later displayed by the pore form [24] , [25] ( Figure S4 ) . In agreement with previous modeling we find that the prepore conformation can be explained solely by a tilt ( ∼40° ) of Domain 4 with respect to the long axis of the molecule [8] ( Figure 7B , C ) . Such orientation is supported by the determined solvent exposure of amino acids of Domain 4 [9] ( Figure S5 ) . Based on our previous structural and MD simulations analyses we hypothesize that this minor domain rearrangement may be attributable to the intrinsic flexibility of Domain 4 and deformation of Domain 2 , although the low resolution of the cryo-EM data ( 28 Å ) prevents further interpretation . In addition , the relative orientation of Domain 4 in the prepore conformation is broadly similar to the crystallographic conformation of ILY IB ( Figure 7C ) , and within the range of observed CDC crystallographic conformations ( Figure 7D ) . Stereochemical features of the model are reported in Table 3 . Therefore this indicates that the CDC monomer in the prepore form adopts an arrangement of domains readily accessible in the soluble form , which is conformationally trapped upon oligomerization . To validate this prepore model a suggested site of interaction would be between strand 2 of Domain 2 of one molecule with the TMH1 of the adjacent molecule ( respectively Thr86-Ser88 and Lys201-Asn205 , PFO numbering ) . This could be performed using either disulphide bond formation or short crosslinkers . Disulphide bond formation experiments have successfully characterized Domain 3 oligomeric interactions in PFO [24] . Finally we investigated the conformation of CDCs in the pore form to address how CDCs change conformation particularly with respect to Domain 2 . Our CDC pore model is presented in Figure 8 . Domain 1 can be fitted intact ( see Methods , CCC for the individual domain of 0 . 70 ) into the cryo-EM density in agreement with Tilley et al . [8] . Together with Domains 1 of adjacent subunits the domains exhibit packing similar to the prepore complex . TMH1/2 are entirely restructured from bundles of α-helices to a giant transmembrane β-barrel , concomitantly with the opening of Domain 3 . It has been established that the amino acids forming the β-barrel adopt a novel β-barrel architecture specific to CDCs [24] , [25]: the membrane-embedded β-hairpins adopt a 20° tilt to the axis normal to the membrane ( Figure 8B ) . This departs from a pore model where the β-hairpins are proposed to stand perpendicular ( 0° tilt ) to the membrane surface [8] . The modeled β-barrel is in full agreement with the experimentally established amphipathic pattern of the membrane-spanning PFO β-hairpins [26] , [27] ( CCC of 0 . 57 , Figure S6 ) . The position of the membrane binding Domain 4 is readily identifiable in the density [8] with only the region surrounding the undecapeptide loop inserting into the upper leaflet of the membrane bilayer [28] , [29] . Domain 4 loosely packs against its adjacent counterparts as demonstrated [29] . Its orientation , tilted but not lying on the membrane surface , is further supported by the pattern of solvent exposure of amino acids distributed on the domain's surface [29] ( CCC of 0 . 48 , Figure S7 ) . Taken together , our structural and MD analyses suggest that Domain 2 does not favour the proposed bending [8] . Instead our data suggest that the β–sheet Domain 2 simply untwists and rotates with respect to Domain 4 . Motivated by this finding we modeled the elongated β-sheet without altering its structural integrity ( see Methods ) . We found that Domain 2 can be fitted in the density linked to Domain 1 and the Domain 4 of the adjacent monomer ( clockwise when viewed from the top; Figure 8C ) . This dramatic sideways rotation has not been postulated to date , yet it provides a good fit of the domains within the cryo-EM maps . The Domain 2 region in our model has a CCC of 0 . 52 with no clashes observed between residues , which is excellent for this resolution of cryo-EM data . This is an improvement on the existing unrefined model , which has a CCC of 0 . 45 for the Domain 2 region ( PDB ID 2bk1 ) . Following the flexible fitting step the 38-mer exhibited a best CCC value of 0 . 67 , an improvement on the initial 0 . 62 . In support of this model we investigated the Domain 2 boundaries with Domain 1 and Domain 4 . Our new model preserves the hydrophobic Domain 2/4 interface , with Domain 2 linked to Domain 4 by a glycine linker . There is an introduction of a kink of ∼40° at the Domain 1/2 interface . The Domain 1/2 interface is constituted by three mainchain covalent links and contains no secondary structure elements or specific contacts . In the pore conformation , Domain 2 orientation is at a ∼25° angle to the bilayer surface and extends the range of orientations observed in crystallographic structures ( Figure 8D , Figure S8 ) . Our model also suggests that the orientation of Domain 2 is constrained by the packing of its adjacent counterparts with the possibility of mainchain parallel hydrogen bonds between positions 54–56 and 384–386 ( PFO numbering ) of an adjacent monomer . In analyzing conformations fitted at such resolution ( 29 Å ) it should be noted that the predicted interaction is indicative of the close proximity of individual Domains 2 in the pore form ( Figure S9 ) . Therefore experiments designed to test this hypothesis should take this aspect into account and may include the use of techniques such as FRET , disulfide bond formation and chemical crosslinking . If strands of Domain 2 are close enough to establish parallel hydrogen bonds then the close proximity can be tested by formation of disulfide bonds . Alternatively in the case of less intimate contacts , chemical crosslinking would be more appropriate . We suggest that probing the Domain 2 oligomeric contact is well suited to distinguishing between the ‘buckling’ model and our proposed model of CDC pore assembly . To assess the stability of our pore model the fitted conformation was energy-minimized and subjected to a brief MD simulation in a membrane bilayer environment and free of all constraints . After 15 ns of simulation the assembly reached a plateau at 4 . 4 Å over the last 5 ns ( Figure S10 ) . The general subunits arrangement remained stable with little deviation from the initial conformation . Minor structural deviations included a difference in the orientation of Domain 4 as well as its penetration into the membrane bilayer . Since the details of its contacts with the bilayer are currently lacking slight deviations are not unexpected . We also noted in the monomer situated at the clockwise end of the tetramer an increased divergence of Domain 2 and the transmembrane Domain 3 ( Figure S10 ) . Given their positions at one extremity of the assembly we concluded that this is likely to be attributable to the lack of explicit circular symmetry in our setup ( see Methods ) and the absence of buttressing/specific contacts with the adjacent monomer . We found the MD simulation demonstrates the overall stability of our pore model and reflects the quality of its stereochemistry ( Table 3 ) . This study has also allowed us to map the rigid bodies present in CDCs and their spectrum of flexibility relative to each other . Following this analysis we have revisited both prepore and pore conformations employing available cryo-EM data . Importantly we find the pore conformation can be modeled without potentially energetically costly restructuring of Domain 2 . Instead our modeling suggests that simple domains rotations can account for the well-documented CDC vertical collapse [8]–[10] . In addition , both our prepore and pore conformations define a pathway for the most logical mechanism of pore formation . Only a coordinated vertical collapse together with rotations of Domains 1/3 ( ∼10° ) is compatible with the extent of Domain 2 rotation ( ∼60° ) from a nearly perpendicular to the membrane surface conformation ( prepore ) to nearly parallel ( pore ) . Furthermore , the oligomeric packing and specific contacts established in the prepore form [23] are likely to impose the constraints that result in a downward spiral rotation of Domains 1/3 ( counter-clockwise rotation corresponding to a monomer and vertical collapse; Figure 9A–C ) within the entire oligomeric assembly . This movement defines an unprecedented and orchestrated global motion whereby the prepore transitions to the pore form by rotation of Domain 2 of all subunits , which brings the CDC head domains ( Domains 1/3 ) closer to the membrane surface . Interestingly , a recent study on aerolysin proposed that a swirling-like motion is central to the mechanism of pore formation by this toxin . While aerolysin and the CDCs are not related , the mechanism we propose is somewhat mechanically analogous [30] . Thus , to conclude , we propose that CDCs achieve pore formation by employing large , concerted domains rotations ( schematically summarized in Figure 9D ) . Our work supports a new model of membrane insertion for CDCs in considerable departure from the currently accepted model . This mechanism presents mechanical similarities to other β-pore forming toxins and a new , testable model of pore formation for CDCs .
The identification of common substructures from structural alignments is a powerful approach that allows us to extract the rigid fragments and conformational properties [15] , [31] of the CDC molecule . The method provides a standardized way to identify common structural cores of homologous proteins ( the “sieving” procedure ) [32] , [33] although it cannot unambiguously distinguish between conformational change and structural divergence . It is , however , important to note that analysis of structures of one protein , and/or homologous proteins , determined in different conditions captures the conformational features across a family [15] , [34]–[36] . All structural alignments were performed with Mustang-MR [37] with domains definition as reported by Rossjohn et al . [5] . Examination and analysis were undertaken using Prody 1 . 4 [38] , Pymol 1 . 3 [39] and VMD 1 . 9 [40] . β-sheet twist values are the angles between mainchain vectors of residues in an inter-strand pair and were calculated following Ho and Curmi [41] . Accessible surface areas are as reported by PISA [42] . MD simulations were analyzed with VMD and Prody . PCA were performed with Prody employing 10 K snapshots collected every 10 ps ( Cα coordinates ) . Initial conformations and MD simulations performed are reported in Table 2 . In all cases topologies were built and solvated using teLeap [43] and the Amber ff99SB force field [44] . MD simulations employed truncated octahedron water boxes ( TIP3P , 12 Å padding ) , sodium and chloride ions were added to charge neutrality . The systems were typically comprised of 164 , 000 to 171 , 000 atoms . Temperature was maintained at 300 K using Langevin dynamics with a damping constant of 5 ps−1 . Pressure was maintained at 1 atm with a Nosé-Hoover-Langevin piston and Periodic Boundary Conditions ( PBC ) were used . An integration time step of 2 fs was used , short-range forces and long-range electrostatics were calculated every time step . Non-bonded interactions employed a 10 Å cut-off and long-range electrostatic forces were computed by the particle-mesh Ewald ( PME ) summation method ( grid spacing smaller than 1 Å ) . All systems were subjected to equilibration steps with harmonic restraints first applied to all heavy protein atoms ( 100 ps , 1 fs time step for this step only ) , followed by restraints applied only to mainchain atoms ( 250 ps ) and finally Cα atoms ( 500 ps ) . 100 ns MD simulations were then produced and analyzed after typically removing the initial 2 to 3 ns . All the simulations were conducted with NAMD v2 . 9 [45] . To obtain initial models of the prepore conformation we first performed rigid body docking into the cryo-EM density of representative CDC crystallographic structures in order to leverage the conformational variability highlighted in our analysis . Four crystallographic structures were used: PFO I , PFO IIIA , ILY IA , ILY IB ( cf . Table 1 ) . The PLY sequence ( UniProt ID: Q04IN8 ) was first threaded onto each structural template employing Modeller 2 . 11 [46] and amino acids corresponding to β5 were simply discarded . The coordinates of the loop at positions 95 to 101 ( PFO numbering ) were also discarded as the corresponding amino acids were found to produce steric hindrance upon oligomeric assembly . The coordinates were then docked using Situs [47] into the cryo-EM map reconstruction of the PLY prepore ( EMDB ID: 1106 ) with the density of the membrane discarded for this step only as including the bilayer density produced unrealistic placements . The docked individual subunits displayed CCCs ranging from 0 . 66 to 0 . 76 ( truncated map ) and 0 . 63 to 0 . 71 ( no truncation ) . Secondly , for each template the docked structure with the highest CCC was then replicated with C31 symmetry and the positions and orientations of all subunits refined against the density with Situs . Finally , four consecutive subunits ( referred to as tetramer in the following ) were selected and the missing loop modeled with Modeller , with amino acids corresponding to the β5 region modeled as an α-helix ( see text ) . This led to the production of five initial PLY conformations with a best CCC of 0 . 57 . Each tetramer was then subjected to a step of flexible fitting into the cryo-EM density with C31 symmetry restraints ( see relevant section ) . A total of 12 models were thus produced with an average root mean square deviation ( rmsd ) of 2 . 4±0 . 9 Å and an average CCC of 0 . 60±0 . 01 . Although small differences in orientation of the domains were observed ( as reflected by the rmsd ) all models presented the same domain architecture ( discussed in the text ) . Cα coordinates of the model with optimal CCC ( 0 . 61 ) are provided as Dataset S1 . Domain 1 was initially placed manually in the cryo-EM density ( EMD ID: 1107; C38 symmetry; 29 Å resolution ) and its position refined locally in the presence of symmetric subunits employing Chimera 1 . 8 [48] . Here we modeled Domain 3 as a β-barrel with architecture S = n/2 as detailed in our previous work [25] . Domain 1 orientation was then adjusted to satisfy both a reasonable fit to the density as well as the distance constraints from the four covalent bonds linking Domain 1 and the four β-strands forming the β-barrel ( Domain 3 ) . Before refinement into the cryo-EM density the altitude of Domain 4 was adjusted so as to place the amino acids identified as exposed to solvent and buried in the membrane [29] ( Figure S6 ) . Domain 2 was placed manually in the density without modifications to its internal structure , consistent with the conclusions of our structural analysis . Its initial placement also satisfies the distance constraints imposed by the covalent bonds to Domain 1 and 4 and a reasonable fit to the cryo-EM map . Our pore model was built as a tetramer with C38 symmetry . Initial coordinates of PFOIII-A were employed with the PLY sequence threaded . Initial positions of all domains were adjusted to remove inter-subunit steric clashes . Furthermore , the initial coordinates were perturbed by 1° clockwise and anti-clockwise rotations around the pore axis thus producing three starting points for the flexible fitting step . The best CCC was 0 . 62 for the initial conformation . The three tetramers were subjected to a step of flexible fitting into the cryo-EM density with C38 symmetry ( see relevant section ) . Final average rmsd was 1 . 4±0 . 4 per monomer and the average CCC for the 38-mer assembly was 0 . 66±0 . 01 . Cα coordinates of the model with optimal CCC ( 0 . 67 ) are provided as Dataset S2 . Flexible fitting was performed following the MDFF methodology [49] with NAMD 2 . 9 . Symmetry restraints were employed for the prepore and pore conformation with the corresponding circular symmetries [50] . All MDFF simulations employed the CHARMM36 forcefield [51] in vacuum ( long range interactions were cut off at 12 Å; dielectric constant of 80; 1 fs time-step; 298 K ) . Additional restraints were applied to preserve correct stereochemistry and prevent structural distortions [52] ( secondary structure restraints force constant of 200 kCal mol−1 rad−2 ) . In each MDFF simulation minimization and equilibration steps were as follows: 10 , 000 steps of minimization with non-hydrogen atoms harmonically constrained , 100 , 000 steps of equilibration with protein main-chain constrained and 100 , 000 steps with Cα atoms constrained . Three 5 , 000 , 000 steps MDFF runs were then performed with linearly increasing symmetry restraints to a final force constant of 5 kCal mol−1 Å−2 driving the system to the desired circular symmetry , and each with different grid force scaling parameter ξ = 0 . 2; 0 . 3; 0 . 5 . Convergence was reached in all cases . Each of the three conformations obtained was subjected to 10 , 000 steps of minimization with ξ = 1 . 0 . Therefore each starting conformation produced three final conformations . A mixed square bilayer membrane patch was generated with DMPC and cholesterol ( 50/50 ratio , 1050 molecules each ) with CHARMM-GUI 1 . 4 [53] , [54] and the CHARMM36 forcefield [55] and equilibrated for 4 ns following the CHARMM-GUI provided settings in the presence of a TIP3P water layer of 15 Å thickness , 0 . 15 M sodium/chloride ions and planar constraints with the NAMD 2 . 9 software . The dimensions of the equilibrated bilayer system were 205 Å×205 Å×71 Å . The tetramer pore model was energy-minimized for 2 , 500 steps free of cryo-EM restraints employing the Generalized Born/Solvent Accessible Surface Area implicit solvation [56] . Bilayer and solvent were then added and their height manually adjusted to match the position of the bilayer as judged from the cryo-EM density . Waters and lipids within 1 . 4 Å of the protein assembly were discarded . TIP3P waters were then added to a system of initial dimensions 205 Å×205 Å×175 . 5 Å . Ions were added to 0 . 15 M and system charge neutrality . The system was heated to 300 K and equilibrated in steps for 4 ns , first melting the lipid tails and cholesterol , then the headgroups and solvent and finally smoothly relaxing harmonic restraints on the tetramer . Care was taken to keep water molecules outside of the bilayer in the first steps of equilibration . The system was then simulated for 15 ns in the NPAT ensemble free of constraints with PBC . Temperature was maintained at 300 K using Langevin dynamics with a damping constant of 1 ps−1 . Pressure was maintained at 1 atm with a Nosé-Hoover-Langevin piston . An integration time step of 1 fs was used , short-range forces and long-range electrostatics were calculated every 1 and 2 fs respectively . Non-bonded interactions employed a 12 Å cut-off with a shorter Lennard-Jones switching function ( 11 to 12 Å ) [55] , long-range electrostatic forces were computed by the PME summation method ( grid spacing smaller than 1 Å ) . The final dimensions of the system were 204 Å×204 Å×164 Å ( 709 , 660 atoms ) . Simulations were performed with the Multi-modal Australian ScienceS Imaging and Visualisation Environment ( MASSIVE ) [57] .
|
Pore formation is central to the ability of cholesterol dependent cytolysins ( CDCs ) to act as important bacterial virulence factors . Secreted by numerous pathogens the toxins assemble into a circular ring and then perforate the target membrane to form the largest self-assembling proteinaceous pores known . In this paper we investigated computationally the conformational properties of the CDC molecule and deduced a new structural model of pore formation and membrane insertion that reconciles all experimental data . The mechanism of membrane perforation by CDCs put forward here reveals concerted and unsuspected domains motion of large amplitude , which conflicts with the currently proposed model . The work presented here procures a plausible structural mechanism of CDC oligomeric transition and furthers our understanding of pore formation by these important toxins .
|
[
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Methods"
] |
[
"stereochemistry",
"molecular",
"dynamics",
"molecular",
"conformation",
"chemistry",
"physical",
"sciences",
"computational",
"chemistry"
] |
2014
|
A New Model for Pore Formation by Cholesterol-Dependent Cytolysins
|
The human pathogen Candida albicans can assume either of two distinct cell types , designated “white” and “opaque . ” Each cell type is maintained for many generations; switching between them is rare and stochastic , and occurs without any known changes in the nucleotide sequence of the genome . The two cell types differ dramatically in cell shape , colony appearance , mating competence , and virulence properties . In this work , we investigate the transcriptional circuitry that specifies the two cell types and controls the switching between them . First , we identify two new transcriptional regulators of white-opaque switching , Czf1 and white-opaque regulator 2 ( Wor2 ) . Analysis of a large set of double mutants and ectopic expression strains revealed genetic relationships between CZF1 , WOR2 , and two previously identified regulators of white-opaque switching , WOR1 and EFG1 . Using chromatin immunoprecipitation , we show that Wor1 binds the intergenic regions upstream of the genes encoding three additional transcriptional regulators of white-opaque switching ( CZF1 , EFG1 , and WOR2 ) , and also occupies the promoters of numerous white- and opaque-enriched genes . Based on these interactions , we have placed these four genes in a circuit controlling white-opaque switching whose topology is a network of positive feedback loops , with the master regulator gene WOR1 occupying a central position . Our observations indicate that a key role of the interlocking feedback loop network is to stably maintain each epigenetic state through many cell divisions .
Transcriptional circuits are central to the regulation of many biological processes . Often the logic of the circuit , rather than the nature of its components , makes up its most critical feature . In this paper we describe an interlocking network of positive feedback loops that underlies white-opaque switching in the human fungal pathogen Candida albicans . White-opaque switching is an epigenetic phenomenon , where genetically identical cells can exist in two distinctive cell types , white and opaque [1] . Each cell type is stably inherited for many generations , and switching between the two types of cells occurs stochastically and rarely—roughly one switch in 104 cell divisions [2] . The white form is the default cell type , and we propose that the main purpose of the network of interlocking feedback loops is , once excited , to stably maintain the opaque cell type through many cell divisions . Thus , we propose that the feedback loop network driving opaque formation is activated infrequently , but once activated , it is stably propagated through many cell generations . Despite possessing the same genome , white and opaque cells have many phenotypic differences . Approximately 400 genes are differentially expressed between the two cell types , and the cells differ in their appearance under the microscope and in the color and shape of the colonies they produce on solid media [1 , 3 , 4] . They also differ in their behavior toward other C . albicans cells: opaque cells , but not white cells , are highly competent for mating; they respond to mating pheromone with polarized growth , and they can subsequently undergo cell and nuclear fusion with opaque cells of the opposite mating type [5–8] . Finally , the two types of cells appear to interact differently with their mammalian host , with opaque cells appearing more suited for skin infections , and white cells appearing to fare better in blood stream infections [9 , 10] . Several transcriptional regulators have been identified that play key roles in maintaining the white and opaque cell types , and in controlling the switching between them . Cells of mating type a/α are blocked for white-opaque switching , with all cells remaining locked in the white phase [5] . This block occurs through the action of two homeodomain proteins ( a1 and α2 ) , encoded at the mating type-like a ( MTLa ) and MTLα loci , respectively . The a1 and α2 proteins likely act together as a heterodimer to repress transcription of WOR1 , the product of which is a positive regulator of the opaque state [11–13] . Wor1 is required for establishment and maintenance of the opaque state , and ectopically expressed WOR1 drives cells into the opaque form . In a and α cells ( both of which are permissive for switching ) , WOR1 is expressed at low levels in white cells , but in opaque cells Wor1 activates its own synthesis , and WOR1 expression levels rise dramatically . High levels of Wor1 , produced by this positive feedback loop , are necessary to maintain the opaque state . Finally Efg1 , which has been studied extensively for its role as regulator of the filamentous growth and pathogenesis in a/α ( i . e . , non-switching ) strains of C . albicans , also plays a part in white-opaque switching: in a and α cells ( but not a/α cells ) , cells deleted for EFG1 exist almost exclusively in the opaque state [14 , 15] ( this work ) . Thus EFG1 is needed to stably maintain the white state . In this paper we identify two additional transcription regulators of white-opaque switching , CZF1 and white-opaque regulator 2 ( WOR2 ) . The former has been previously studied as an important regulator of filamentous growth in a/α ( non-switching ) cells [16] , but a role in white-opaque switching has not been previously described . WOR2 has not been previously described in any context , and we named the gene WOR2 based on its key role white-opaque switching , as described in this paper . In order to understand the genetic relationships among WOR1 , EFG1 , CZF1 , and WOR2 , we constructed a large set of single and double mutants and analyzed them for white-opaque switching . We also ectopically expressed these regulators in mutant strains in various combinations and monitored their effects on switching and maintenance of the white and opaque states . Finally , we carried out chromatin immunoprecipitation ( ChIP ) experiments to establish direct regulatory connections between the central regulator , Wor1 , and the other targets . We found that Wor1 binds upstream of: ( 1 ) all four transcriptional regulators investigated in this paper ( WOR1 , CZF1 , WOR2 , EFG1 ) ; ( 2 ) genes whose transcription is regulated by the white-opaque switch; and ( 3 ) a large number of genes that are not differentially transcribed during the white-opaque switch , suggesting an additional , previously unrecognized component of white-opaque switching , one that may require additional environmental inputs to fully reveal . Based on the combined results of these experiments , we have placed MTLa1 , MTLα2 , WOR1 , CZF1 , WOR2 , and EFG1 into a single genetic circuit regulating the white-opaque switch . This circuit is formed from a network of interlocking positive feedback loops , and we believe that this network can account for the stability of the white and opaque states through many cell generations .
One of the most striking characteristics of the white-opaque switch is the large number of genes that are differentially regulated between the two types of cells . Approximately 400 genes have altered transcription; roughly half are up-regulated in the white phase , with the remaining half up-regulated in the opaque phase [3 , 4] . Several of these regulated genes encode transcriptional regulators , as predicted by the presence of a sequence-specific DNA-binding motif encoded in the gene . We tested a set of opaque-enriched transcription factors for possible roles in regulating the white-opaque switch . Opaque-enriched genes were previously identified through microarray analyses that compared the gene expression profiles of isogenic white and opaque cells [3 , 4] . Genes up-regulated in opaque cells were searched for homology to transcriptional regulatory proteins using BLAST searches ( http://www . ncbi . nlm . nih . gov/BLAST/ ) and Pfam motif searches ( http://www . sanger . ac . uk/Software/Pfam/ ) . From a set of 237 opaque-enriched genes , we chose to study six genes encoding putative transcriptional regulators: CZF1 , WOR2 , HAP3 , orf19 . 4972 , CSR1 , and PHO23 . Both CZF1 and WOR2 are predicted to contain a Zn ( 2 ) -Cys ( 6 ) zinc cluster motif , a known DNA-binding domain in fungal transcriptional regulators [17]; indeed WOR2 had been provisionally named ZCF33 to indicate it was the 33rd protein annotated with this zinc cluster motif . HAP3 is predicted to contain a motif similar to the CCAAT-binding factor . The genes CSR1 and orf19 . 4972 are predicted to each contain multiple C2H2 zinc fingers , a well-characterized DNA-binding domain . Because chromatin structure has been proposed to play a role in regulating the white-opaque switch [18] , we also chose to investigate the opaque-enriched transcript PHO23 , which encodes a protein containing a PHD domain and is predicted to be a part of the RPD3 histone deacetylase complex . For each of these candidate genes , we attempted to make homozygous deletion mutants in a white strain that is mating type a , and is thus permissive for switching to the opaque cell type ( C . albicans is diploid , and it is therefore necessary to knock out two copies of each gene ) . Multiple independent deletion mutants of each target gene were made from independent heterozygous mutants . Despite numerous attempts , we were unable to create a homozygous knockout mutant of the CSR1 gene and did not study CSR1 further . White-opaque switching can occur in strains that are mating type a or α but not a/α . Most of the work presented in this paper was performed in a strains , but we know that Wor1 is also required for opaque formation in α strains ( unpublished data and [13] ) , suggesting that white-opaque switching is controlled the same way in a and α strains . Consistent with this idea , a large set of microarray data indicates that the set of genes regulated by white-opaque switching is virtually identical in a and α cells [3] . For the five remaining candidates , we performed quantitative white-to-opaque switching assays as described previously [5] on at least two independent deletion mutants for each candidate gene , and multiple experiments were performed for each mutant ( Table 1 ) . As shown in Table 1 , two mutants , the czf1Δ/czf1Δ and wor2Δ/wor2Δ knockouts , had a dramatic effect on white-opaque switching , forming opaque colonies much less frequently than did otherwise isogenic wild type ( WT ) a strains . The CZF1 deletion strain formed opaque sectors and colonies ∼50-fold less frequently than WT a strains . The wor2Δ/wor2Δ mutant was never observed to form opaque colonies , representing a switching frequency at least 180-fold below that of the parent strain . Due to the key role this gene has in white-opaque switching , as described in this paper , we named the gene WOR2 . These results implicate both CZF1 and WOR2 in the white-opaque switch; formally , they function as activators of the opaque state . To verify that the defects in white-opaque switching were attributable to the disrupted genes , we complemented the czf1Δ/czf1Δ and wor2Δ/wor2Δ deletion mutants . Ectopic expression constructs , controlled by the MET3 promoter were introduced into the RP10 locus , as described previously [19] . Both the czf1Δ/czf1Δ pMET3-CZF1 and wor2Δ/wor2Δ pMET3-WOR2 strains were able to form opaque colonies when the MET3 promoter was induced . However , when an empty vector was introduced , or the strains were grown on media that repressed the MET3 promoter ( and thus the only copy of CZF1 or WOR2 , respectively ) , the strains remained white . These results confirmed that the loss of CZF1 or WOR2 drastically reduces the ability for the strains to grow as opaque cells . The deletion mutants lacking either orf19 . 4972 or HAP3 formed opaque colonies at frequencies comparable to WT a strains and were not studied further . The pho23Δ/pho23Δ mutant switched to the opaque phase approximately six times as frequently as the WT control . If Pho23 works with Rpd3 in C . albicans , as is predicted based on homology in Saccharomyces cerevisiae , this result is consistent with a previous finding that rpd3Δ/rpd3Δ mutants have an increased frequency of interconversion between the white and opaque phases [18] . We did not study Pho23 further , because of its relatively small affect on switching frequencies , and because these effects could well be indirect: in S . cerevisiae , deletion of RPD3 affects transcription levels of approximately 13% of the genome [20] . We next expressed CZF1 and WOR2 ectopically in white cells to test whether either could drive white cells to the opaque form . All ectopic expression constructs described in this study were controlled by the MET3 promoter integrated at the RP10 locus , as previously described [19] . To test if ectopic expression of a given gene causes white-opaque switching , white strains were streaked from frozen stock onto repressing media and grown at room temperature for 1 wk . The strains were then plated for single colonies onto inducing media or repressing media , as a control , and grown for 1 wk at room temperature . The control a strain , with an empty vector ( pCaEXP ) inserted into the RP10 locus , switched to the opaque phase at the typical low frequency , producing opaque sectors in approximately 0 . 5% of the colonies on both media conditions ( Table 2 ) , indicating that the media conditions used to control the MET3 promoter do not significantly influence the frequency of white-opaque switching . We found that ectopic expression of CZF1 in WT a cells led to a mass conversion to the opaque phase ( Table 2 ) , but only when the MET3 promoter was induced . In contrast , expression of a pMET3-WOR2 construct did not drive the white-to-opaque switching; the cells remained white , based on colony appearance ( Table 2 ) and cell shape ( unpublished data ) . We know that the pMET3-WOR2 construct is functional from the complementation studies described earlier , thus this result indicates that the ectopic expression of WOR2 , at least at the level driven by the MET3 promoter , is not sufficient to drive opaque formation in an otherwise WT a strain . Previous work on the regulation of white-opaque switching identified WOR1 as a master regulator of the white-opaque switch [11–13] . In the next set of experiments , we tested the genetic interactions between WOR1 , CZF1 , and WOR2 in order to understand how they work together to regulate the switch . As is the case for CZF1 and WOR2 , deletion of WOR1 drastically reduces the frequency of opaque formation . Like CZF1 , ectopic expression of WOR1 causes mass conversion to the opaque phase in otherwise WT a cells . We first expressed WOR1 ectopically in a czf1Δ/czf1Δ or wor2Δ/wor2Δ a-cell strain and observed the effects on white-opaque switching . We found that when WOR1 was ectopically expressed in white czf1Δ/czf1Δ mutants , most of the colonies grew in the opaque phase or had opaque sectors ( Table 2 ) , although the colonies had a slightly rougher texture than did conventional opaque colonies on inducing media ( unpublished data ) . Inspection of the cells from the opaque colonies revealed elongated cells , typical of opaque cells ( Figure 1 ) . In a control experiment , a czf1Δ/czf1Δ mutant with pCaEXP , an expression vector lacking WOR1 , was not converted into opaque cells ( Table 2 ) . When we expressed WOR1 ectopically in a white wor2Δ/wor2Δ a strain , we observed that all of the colonies contained cells that had switched to the opaque phase ( Figure 1 ) , usually in the form of opaque sectors , though the opaque sectors were slightly lighter in color than those of normal opaques ( Table 2 ) . This strain was never observed to form opaque cells when grown on media that repressed expression of the ectopic WOR1 . The wor2Δ/wor2Δ strain with pCaEXP , an empty expression vector , also appeared locked in the white phase , whether it was grown on the repressing or inducing media ( Table 2 ) . Next , we tested the effects of ectopic expression of CZF1 in a wor1Δ/wor1Δ a strain . We found that these strains remained locked in the white phase ( Table 2 ) ; they were indistinguishable from a wor1Δ/wor1Δ mutant . Finally , we tested the ectopic expression of WOR2 in a wor1Δ/wor1Δ a strain , and we found no change in the switching frequency , as compared to a wor1Δ/wor1Δ mutant ( Table 2 ) . This result was expected , given that induction of the WOR2 ectopic construct had no effect in a WT background . Taken together , these results indicate that CZF1 and WOR2 function upstream of WOR1; thus ectopic expression of WOR1 suffices for opaque cell formation whether or not WOR2 and CZF1 are present . However , the converse is not true: deletion of WOR1 cannot be overcome by ectopic expression of CZF1 . Unlike wor1Δ/wor1Δ and wor2Δ/wor2Δ mutants , czf1Δ/czf1Δ mutants do form opaque colonies , albeit infrequently . As described above , ectopic CZF1 expression can induce a switch to the opaque state . To clarify CZF1's role in white-opaque switching , we examined switching in the reverse direction , where opaque cells switch to white cells . When opaque isolates of WT a strains were replated on repressing media , about 16% of the colonies switched back to the white form ( Table 1 ) . The rare opaque isolates of czf1Δ/czf1Δ a strains were nearly as stable as WT opaques; upon replating , 23% of the colonies contained white cells . Because opaque isolates in czf1Δ/czf1Δ strains are rare , we sought to create more opaque czf1Δ/czf1Δ isolates in order to test the stability of the opaque cells lacking Czf1 . To do this , we used the pMET3-WOR1 ectopic expression construct to drive czf1Δ/czf1Δ strains to the opaque state , as described above . When the pMET3-WOR1 construct was subsequently repressed in the czf1Δ/czf1Δ opaque a strains , at least 92% of the colonies remained opaque ( Table 3 ) . Similarly , a pulse of pMET3-WOR1 in WT white a cells is sufficient to generate stable opaque populations; the ectopic Wor1 expression can be repressed and the strains will largely continue to grow in the opaque phase ( Table 3 ) [11] . These data indicate that , although its presence is important to form opaque cells , Czf1 contributes minimally to the stability ( that is , the heritability ) of the opaque state , once it has been established . In parallel with the studies described above , we examined WOR2's role in maintaining the heritability of the opaque state . As described , when WOR1 was ectopically expressed in wor2Δ/wor2Δ mutants , opaque cells formed . When the pMET3-WOR1 construct was then repressed in these cells ( Table 3 ) , the majority of the cells reverted to the white form ( Table 3 ) . In contrast , WOR2/WOR2 control strains remained in the opaque form for many generations after the pulse of WOR1 expression . These results indicate that Wor2 contributes greatly to the stability of the opaque state , once it has been formed . Thus far , we have only considered the role of the opaque-enriched transcription factors WOR1 , CZF1 , and WOR2 in the regulation of the white-opaque switch . However , a fourth regulator , EFG1 , which is up-regulated in white cells , is known to participate in white-opaque switching [14 , 15] . Experiments reported by Sonneborn et al . [14] suggested that depletion of Efg1 induced the formation of opaque cells in some a/α strain backgrounds , but not in others . To clarify these results , we constructed new isogenic homozygous efg1Δ/efg1Δ mutants in a or a/α strains . In the mating type a strain , we found the efg1Δ/efg1Δ mutation caused a majority of the population to switch to the opaque phase; over 98% of the colonies contained opaque sectors ( Table 4 ) , with many colonies showing multiple sectors . We also observed a small number of entirely white colonies , indicating that EFG1 is not strictly necessary for growth in the white phase . We also examined the opaque-to-white switching frequency in efg1Δ/efg1Δ mutants; we found that they switched to the white phase ∼80 times less frequently than WT a strains ( Table 4 ) . Thus , deletion of EFG1 dramatically increased the likelihood the cells will grow in the opaque phase , confirming previous studies in WO-1 , an α strain [15] . In contrast to the previous reports , we never observed opaque colonies or sectors in the efg1Δ/efg1Δ mutant in an a/α strain , despite observing over 3 , 200 colonies ( unpublished data ) . We obtained the previously published efg1Δ/PCKpr-EFG1 a/α mutant that showed opaque cell formation when remaining allele of EFG1 was repressed [14] . Using PCR to amplify the MTLa1 and MTLα2 genes , we determined that this mutant was an a strain , likely due to spontaneous loss of one copy of Chromosome 5 , which carries the MTL locus ( Figure 2 ) . Additionally , each of the efg1Δ/efg1Δ mutants that were locked in the white phase was confirmed to be an a/α strains ( Figure 2 ) . These results indicate that the loss of Efg1 from a cells causes massive conversion to the opaque state , but this conversion is blocked in a/α cells . Thus Efg1 functions upstream of the a1-α2 block of white-opaque switching . To understand the genetic interplay between EFG1 , CZF1 , and WOR2 , we created strains that lacked the white-enriched regulator EFG1 and each of the opaque-enriched regulators , CZF1 or WOR2 . These double homozygous knockouts were then tested in quantitative switching assays and monitored for the frequency of forming white , opaque , and sectored colonies . Nearly all colonies of the efg1Δ/efg1Δ czf1Δ/czf1Δ mutant were in the opaque phase or contained opaque sectors ( Table 4 ) , reflecting the phenotype of the efg1Δ/efg1Δ mutant . When the opaque colonies isolated from efg1Δ/efg1Δ czf1Δ/czf1Δ were replated to test the heritability of the state , only 0 . 08% of the colonies returned to the white state , in comparison with the normal opaque-to-white switching frequency , where ∼16% of the colonies are white or have white sectors . Thus , the efg1Δ/efg1Δ czf1Δ/czf1Δ opaque cells are approximately 200-fold more stable than WT opaque cells , a stability similar to that of efg1Δ/efg1Δ a mutants ( Table 4 ) . Thus , in both forward and reverse switching frequency , the efg1Δ/efg1Δ czf1Δ/czf1Δ double mutants closely resembled the efg1Δ/efg1Δ single mutant . We also examined the switching behavior of an efg1Δ/efg1Δ wor2Δ/wor2Δ mutant . In this strain , white colonies accounted for ∼99% of the total colonies seen , reflecting the phenotype of the wor2Δ/wor2Δ mutant . We also tested the stability of these rare opaque colonies that were formed by the efg1Δ/efg1Δ wor2Δ/wor2Δ mutant . When replated , these opaque cells proved to be highly unstable; over 98% of the colonies were white or contained white sectors . Thus , in both forward and reverse switching , the efg1Δ/efg1Δ wor2Δ/wor2Δ double mutant resembled the wor2Δ/wor2Δ mutant . The genetic epistasis data presented above places WOR1 at the center of white-opaque regulation; formally , it is the most “downstream” regulator of opaque formation , as its deletion blocks white-opaque switching in all contexts tested . Moreover , ectopic WOR1 expression suffices to switch white cells to opaque cells when any of the other opaque-enriched transcription factors are deleted . Previous work indicated that Wor1 expression is maintained through an auto-stimulatory positive feedback loop , mediated by Wor1 binding at its own promoter [11] . Expression of the genes encoding Czf1 , Wor2 , and Efg1 are all regulated by white-opaque switching; in the opaque form , CZF1 and WOR2 are up-regulated , and EFG1 is down-regulated relative to the white form . Thus , in a formal sense all three are regulated by Wor1 . To test whether Wor1 directly regulates CZF1 , WOR2 , and EFG1 , we performed ChIP using an affinity purified antibody ( α-Wor1Nterm ) , raised against a peptide near the N terminus of Wor1 . These ChIPs were analyzed genome-wide using microarrays ( ChIP-chip ) : the precipitated DNA was amplified , fluorescently labeled , and competitively hybridized against genomic DNA ( input DNA ) on custom DNA tiling microarrays containing 60-mer oligonucleotides tiled at 80 bp intervals across the entire C . albicans genome . Two microarrays were hybridized with DNA from two separate immunoprecipitations ( IPs ) of an opaque WT a strain; a single ChIP was performed in a wor1Δ/wor1Δ ( white ) a strain as a control . We examined the Wor1 ChIP data using previously published software that implements a statistical model and integrates data from several neighboring spots along chromosomes to identify IP enrichment peaks [21] . Using standard parameters , this procedure identified 206 peaks of Wor1 enrichment across the genome in opaque cells . These peaks of Wor1 enrichment were confirmed by visual inspection of ChIP-chip data plotted along chromosomes . Of these peaks , 25 also appeared in the ChIP-chip of a wor1Δ/wor1Δ strain , and likely represent cross reactivity or particularly “sticky” proteins; they were removed from further analysis , leaving a set of 181 peaks of Wor1 enrichment . In parallel , we performed a series of ChIP-chip experiments using a different antibody against Wor1 ( raised against a peptide at the C terminus of the Wor1 protein , α-Wor1Cterm ) and identified 122 peaks of Wor1 enrichment in opaque strains that were not detected in the wor1Δ/wor1Δ control strain . By comparing the sets of peaks of Wor1 enrichment identified using both antibodies , we found that 112 peaks are enriched for Wor1 in opaque cells ( but not wor1Δ/wor1Δ strains ) using both antibodies . Thus , 112/122 ( 92% ) of the peaks identified using the α-Wor1Cterm were also found using the N-terminal antibody , indicating the set of peaks identified with α-Wor1Cterm is almost entirely a subset of α-Wor1Nterm ChIP-chip data . We found that 112/181 ( 62% ) of the targets identified in using the α-Wor1Nterm antibody were also detected using the α-Wor1Cterm in opaque cells . Because the experiments using α-Wor1Nterm exhibited very little cross-reactivity in wor1Δ/wor1Δ strains and virtually encompassed the set found using α-Wor1Cterm , we chose the targets identified in the α-Wor1Nterm ChIP-chip as our set of high-confidence Wor1 targets for further analysis . With 181 peaks identified as high confidence Wor1 targets using the α-Wor1Nterm antibody , we turned to the question of identifying the genes potentially regulated by Wor1 . We chose to limit the set to the 170 peaks positioned in intergenic regions upstream of at least one open reading frame ( ORF ) ; this eliminated three peaks positioned within ORFs and eight peaks positioned between convergent ORFs . Because some of the 170 peaks of Wor1 enrichment lay in the intergenic region of divergently transcribed genes , there are 221 genes potentially regulated by Wor1 ( Table S1 ) . We found clear Wor1 enrichment at the intergenic regions immediately upstream of the CZF1 , WOR2 , and EFG1 coding sequences ( Figure 3 ) . In the ∼7 . 6 kb of intergenic sequence upstream of CZF1 , we found segments that were enriched up to 20-fold for Wor1 , as compared to a wor1Δ/wor1Δ control strain . Upstream of the WOR2 coding sequence , we found segments enriched up to 11-fold . Wor1 was enriched up to ∼12-fold in the 10 . 1 kb upstream of EFG1 . We also verified that Wor1 was found upstream of the WOR1 gene using the tiling arrays , showing enrichment up to ∼80-fold in the opaque cells , as compared to the control wor1Δ/wor1Δ strain ( Figure 3 ) [11] . These results confirm that Wor1 is present at its own promoter in opaque cells , and reveal that Wor1 is also present at the promoters of CZF1 , WOR2 , and EFG1 in opaque a cells . We also compared the set of 221 genes potentially regulated directly by Wor1 to the set of genes differentially transcribed between white and opaque cells [3] . We found Wor1 enrichment at the intergenic regions upstream of 38 opaque-enriched genes and 20 white-enriched genes ( Table S1 ) , suggesting that Wor1 directly controls expression of approximately 15% of the genes regulated by white-opaque switching . These results also suggest that Wor1 may function in opaque cells as both a transcriptional repressor and as an activator . Though there is some ambiguity in ascribing Wor1 regulation at divergently transcribed genes , we estimate that Wor1 also binds more than 100 genes that have not been previously identified as white- or opaque-enriched by transcriptional profiling . As described above , Wor1 protein is bound at the intergenic DNA upstream of the genes WOR1 , EFG1 , CZF1 , and WOR2 . Each of these genes has a remarkably long upstream region of DNA ( at least 7 kb in each case ) , and Wor1 appears to be bound at multiple positions along these regions . From the ChIP-chip experiments , we found that occupancy of large intergenic regions is a general characteristic of Wor1 . Analysis of the intergenic regions at all 6 , 077 gene promoters in the C . albicans genome ( excluding those at telomeres ) revealed a median promoter length of 623 bp , whereas the median promoter length of the 181 gene promoters bound by Wor1 was 3 , 390 bp ( unpublished data ) . This preference is especially pronounced when considering the intergenic regions over 10 kb; Wor1 enrichment was seen at 12 of 19 of these intergenic regions . Intriguingly , over half ( seven of 12 ) of these intergenic regions lie upstream of genes encoding sequence-specific DNA binding proteins—WOR1 , RFG1 , TCC1 , WOR2 , ZCF37 , EFG1 , and RME1—suggesting Wor1 may exert much of its control over the white-opaque switch indirectly through other transcriptional regulators .
In this paper , we have dissected the genetic circuitry controlling white-opaque switching in the fungal pathogen C . albicans . White-opaque switching is an epigenetic change between two distinct types of cells , both containing the same genome . The white-opaque switch is crucial for many aspects of C . albicans biology , including interactions with other C . albicans cells ( pheromone sensing and mating ) and interactions with the host ( opportunistic pathogenesis ) . Our results are summarized in Figures 4 and 5 , where the circuitry controlling this switch is diagrammed . This network of positive feedback loops is responsible for the heritability of each state , as well as the frequency of switching between them , and we propose that the structure of this network makes an important contribution to the biology of white-opaque switching . The default state can be considered the white cell type: most clinical isolates of C . albicans are a/α cells , and they are locked in the white state through a1-α2 repression of WOR1 ( Figure 5A ) . However , even in a and α cells , which are permissive for white-opaque switching , the white cell type still seems to be the default , in that white cells are generally more stable than opaque cells ( Figure 5B ) . For example , opaque cells at 24 °C are stable for many generations , but above 30 °C they become unstable and rapidly switch back en masse to the white form , which is stable under these conditions [1 , 2] . There are no known environmental conditions that comparably destabilize the white form . In our model , the opaque form is generated when the series of positive feedback loops shown in Figure 4 become excited ( Figure 5C ) . Thus , in opaque cells , WOR1 likely directly induces CZF1 and WOR2 expression , and in turn , CZF1 and WOR2 both activate WOR1 . CZF1 does this by repressing a repressor of the opaque state ( EFG1 ) , the net effect being a positive feedback loop . The multiple feedback loops observed in the opaque state are reminiscent of those seen in differentiated animal cells , such as those of the Drosophila eye ( [22] , reviewed in [23] ) and the mammalian myoblast ( [24] , reviewed in [25] ) . A series of such feedback loops ( as opposed to a single loop ) buffers the circuit against transient fluctuations in any single regulatory protein and therefore provides additional stability to the excited form of the circuit . In addition , the nature of the circuit probably defines the switching frequency . For example , deletion of CZF1 decreases the white-to-opaque switching frequency by approximately 10-fold , but has little effect on the backward switching rate . Thus , the primary role of CZF1 seems to be in modulating the switching frequency; in contrast , WOR1 and WOR2 are both required to maintain the opaque state; thus their roles are more integral to the switch itself . Although the overall logic of the circuit shown in Figure 4 can explain many features of white-opaque switching , there appear to be several unusual features of the circuit components themselves that likely also play important roles in white-opaque switching . For example , our ChIP-chip experiments revealed that Wor1 binding shows a bias toward genes with unusually long upstream intergenic regions—as defined by the distance from the 5′ end of the ORF to the next annotated coding region . This observation suggests that these genes bound by Wor1 , which include the four encoding transcriptional regulators that form the interlocking feedback loops ( WOR1 , EFG1 , CZF1 , and WOR2 ) are also controlled by a number of other transcriptional regulators . It is known that the frequency of white-opaque switching can be influenced by environmental cues ( e . g . , temperature and oxidative stress ) [1 , 26] , and it seems plausible that different rates of switching could be “set” by individually adjusting the levels of the regulatory proteins that make up the circuit . For example , since deletion of CZF1 reduces the frequency of white-to-opaque switching 10-fold , regulation of the level of Czf1 by environmental signals could directly control the “forward” switching rate . Another unusual feature of the circuit concerns the wide distribution of Wor1 over much of the upstream regions of WOR1 , EFG1 , CZF1 , and WOR2 ( Figure 3 ) , suggesting a highly cooperative transcriptional response to the intracellular levels of WOR1 . This , combined with the interlocking positive feedback loops , could be responsible for the switch-like behavior of the system , specifically the failure to readily observe a cell type intermediate between white and opaque in wild-type switching strains . The switch from the white to the opaque form growth alters transcription of approximately 400 genes . We know that the master regulator Wor1 ultimately controls all of these genes , since deletion of WOR1 locks cells in the white form , and ectopic expression of WOR1 converts white cells en masse to opaque cells . Our ChIP-chip analysis revealed that Wor1 directly regulates approximately 15% of this gene set ( 20 white-enriched genes and 38 opaque-enriched genes ) . Since Wor1 is also bound upstream of CZF1 , EFG1 , WOR2 , and 20 additional transcriptional regulators ( see Table S1 ) , it seems likely that much of white-opaque switching program is regulated indirectly by Wor1 through its effects on other transcriptional regulators . An unexpected outcome of the Wor1 ChIP-chip experiments was the presence of Wor1 at a large number of genes that were not identified as white- or opaque-enriched in previous microarray analyses [3] . There are several explanations for this observation . First , Wor1 could control these genes in both white and opaque cells , with their transcription being unaffected by the white-to-opaque transition . We think this explanation is unlikely because Wor1 is up-regulated 45-fold in opaque cells [3] , and it seems improbable that this change could have no impact on expression of target genes . To test this idea directly , we performed a Wor1 ChIP-chip experiment in white a cells , and found that Wor1 is not bound at any of these target genes ( unpublished data ) . A second possibility , one that we favor , is that Wor1 may occupy the promoters of these 100 genes in opaque cells , preparing their expression to respond to unknown environmental signals , perhaps those generated by the host . According to this idea , the standard laboratory conditions used for transcriptional profiling would not have included the necessary environmental stimuli , and thus these genes would not have been identified as regulated by the white-opaque switch . This idea suggests there are additional aspects to white-opaque switching which have not been previously recognized . Finally we note that white-opaque switching does not appear to be a general feature of fungi , even those that are closely related to C . albicans . Indeed it may have arisen during the long association of C . albicans with its warm-blooded hosts . The evolution of a complex circuit composed of interlocking feedback loops is relatively simple to imagine , as it could occur stepwise simply through the acquisitions of cis-acting sequences in genes for transcriptional regulators used for other purposes in the cell . We note that Czf1 also relieves Efg1-mediated repression of hyphal growth under embedded conditions [27] , and this genetic relationship has been maintained in the regulation of the white-opaque switch . Thus EFG1 and CZF1 have other key functions in the cell—even in cells that are genetically blocked for white-opaque switching—and their involvement in white-opaque switching could well be a recent adaptation , functioning to modulate the stability of the two states and the frequency of switching between them . The independent evolution of interlocking transcriptional feedback loops in a variety of distinct biological contexts ( white-opaque switching in C . albicans , eye development in flies , and muscle development in mammals , for example ) suggests they are particularly effective ways of providing , from the same genome , distinctive cell types that can be stably propagated for many generations .
Standard laboratory media have been described previously [28] . Synthetic complete media , supplemented with 2% glucose and 100 μg/ml uridine ( SCD+Urd ) was used to maintain strains in the white and opaque phases at room temperature . For ectopic expression experiments , cells were grown on inducing media ( SCD−Met−Cys+Urd ) or repressing media ( SCD+Met+Cys+Urd ) to control expression of the MET3 promoter , as described previously [11 , 19] . The plasmid containing the pMET3-WOR1 construct ( pRZ25 ) has been described before [11] . To make the pMET3-WOR2 and pMET3-CZF1 constructs , the WOR2 or CZF1 ORFs was PCR-amplified from SC5314 genomic DNA using primers containing BamHI and SphI restriction sites , and cloned into a BamHI/SphI-digested pCaEXP , to create the plasmids pAJ2230 and pAJ2231 , respectively . All strains were derived from SC5314 . EFG1 , CZF1 , or WOR2 was deleted using a modified Ura-blaster protocol [29] . In short , the recyclable URA3-dpl200 marker was PCR-amplified from pDDB57 using long oligonucleotides identical to the sequence immediately flanking each ORF targeted for deletion . The deletion construct was transformed into CHY439 ( α1Δα2Δ , Ura− ) or CAI4 ( a/α , Ura− ) and transformants were selected on SD−Ura media . 5-Fluoro-orotic acid was used to counterselect against URA3 marker , and the resulting Ura− isolates were used for subsequent rounds of gene deletion or to create the ectopic expression strains . For each knockout target , at least two homozygous deletion mutants were created from independent heterozygous mutants . When creating double mutants , CZF1 and WOR2 were each deleted in an efg1Δ/efg1Δ ( α1Δα2Δ , Ura− ) mutant . In the case of the efg1Δ/efg1Δ wor2Δ/wor2Δ mutant , two independent double mutants were created from two independent efg1Δ/efg1Δ homozygous deletion mutants . Each WOR1 allele was deleted from the strain SNY78 ( a/α , His− , Leu− , Ura− ) using fusion knockout constructs described previously [11] . The resulting strain was grown on sorbose-containing media to generate a/a strains ( see [6] and references therein ) , creating the wor1Δ/wor1Δ ( a/a , Ura− ) strain . Ectopic expression constructs pAJ2230 , pAJ2231 , pRZ25 ( described above ) , or pCaEXP ( empty control vector [19] ) were linearized to direct integration to the RP10 locus and transformed into Ura− isolates of WT , wor2Δ/wor2Δ , czf1Δ/czf1Δ , or wor1Δ/wor1Δ strains . To create the duplicate ectopic expression strains listed in Table S2 , ectopic expression constructs were introduced into independent wor1Δ/wor1Δ or wor2Δ/wor2Δ strains . The czf1Δ/czf1Δ ( Ura− ) strains used to create the czf1Δ/czf1Δ + pMET3-WOR1 ectopic expression strains are different Ura− loopout isolates generated by 5-fluro-orotic acid counter-selection of the same czf1Δ/czf1Δ ( Ura+ ) strain . The czf1Δ/czf1Δ + pMET3-CZF1 complementation strains were made from the same czf1Δ/czf1Δ knockout strain . Switching frequencies between the white and opaque phases were determined in plate-based assays , as described previously , with modifications [5] . Strains were streaked from frozen stock onto SCD+Urd and grown at RT for 5–7 d . For each strain , at least five entirely white colonies were resuspended into dH2O , diluted , and plated for single colonies on SCD+Urd . After growth at RT for 1 wk , we examined the colonies and counted the number of switch events ( as evidenced by the presence of opaque sectors , or entirely opaque colonies ) . The same process was used to assess opaque-to-white switching , but the original frozen stocks contained opaque isolates of each strain , and we monitored switching by the presence of white sectors or entirely white colonies . The data shown in Tables 1 and 4 were taken from the same representative experiment and only tested one strain of each genotype . In repetitions of the switching assays ( unpublished data ) , multiple independent deletion mutants of each genotype were tested and yielded results similar to those shown in Tables 1 and 4 . Switching assays in strains containing the pMET3 ectopic expression constructs were performed as described in [11] , with modifications . In short , to test if ectopic expression can drive opaque formation , white strains were streaked from frozen stock onto repressing media at RT for 5 d . At least five fully white colonies were replated for single colonies on inducing media ( or repressing , as control ) . After growth at RT for 1 wk , colony phenotypes were recorded . Colonies were resuspended in sterile water and cells were examined by using differential interference contrast microscopy on an Axiovert 200M microscope ( Carl Zeiss , http://www . zeiss . de/ ) . All experimental strains , excepting the wor1Δ/wor1Δ + pMET3-WOR2 strains , were tested in at least two repetitions of the switching assay . Data shown in Table 2 are from a single representative experiment , and each strain listed is an independent ectopic expression mutant . To test if the resulting colony phenotypes were stable after the ectopic expression was repressed , opaque strains ( formed by induction of the ectopic expression constructs ) were streaked from frozen stock onto inducing media at room temperature . At least five opaque colonies were resuspended in sterile dH2O and replated onto repressing media ( or inducing , as control ) and grown at RT for 1 wk . Colony phenotypes were recorded . Two independently derived strains were tested for each ectopic expression scenario , and experiments were performed at least twice . Data shown in Table 3 are from a single representative experiment , and each strain listed is an independent ectopic expression mutant . To determine the mating type of C . albicans strains , we PCR amplified the a and α alleles of each gene located within the MTL locus ( PAP , OBP , PIK , MTLa1 , MTLa2 , MTLα1 , MTLα2 ) [30] . PCR products for every a allele were seen in all strains tested; products for each α allele were seen in all strains except SS4 ( unpublished data ) . PCR products for the MTLa1 and MTLα2 genes are shown in Figure 2 . Overnight cultures ( 200 ml ) were grown in SCD+Urd for approximately 16 h at 25 °C to an OD600 of 0 . 4 . Cells were formaldehyde cross-linked by adding formaldehyde ( 37% ) to a 1% final concentration . Treated cultures were mixed by shaking and incubated for 15 min at room temperature . 2 . 5 M glycine was added to a final concentration of 125 mM , and treated cultures were mixed and incubated 5 min at room temperature . Cells were pelleted at 3 , 000 g for 5 min at 4 °C and washed twice with 100 ml of 4 °C TBS ( 20 mM TrisHCl [pH 7 . 6] , 150 mM NaCl ) . Spheroplasting and ChIP were carried out as previously described , with modifications [11 , 31] . Cell pellets were resuspended in 39 ml of Buffer Z ( 1 M sorbitol , 50 mM Tris-Cl [pH 7 . 4] ) , 28 μl of β-ME was added ( 14 . 3 M , final concentration 10 mM ) , and cells were vortexed . 20 μl of lyticase ( Sigma , MO , United States ) solution ( 2 mg/ml in Buffer Z ) was added , and cell suspensions were incubated 15 min at 30 °C . Spheroplasted cells were then spun at 3 , 000 g , for 10 min , at 4 °C and resuspended in 500 μl of 4 °C lysis buffer ( 50 mM HEPES-KOH [pH 7 . 5] , 140 mM NaCl , 1 mM EDTA , 1% Triton X-100 , 0 . 1% sodium deoxycholate ) with protease inhibitors . All subsequent ChIP and wash steps were done at 4 °C . DNA was sheared by sonication ten times for 10 s at power setting 2 on a Branson 450 sonicator ( http://www . bransonultrasonics . com/ ) , incubating on ice for 2 min between sonication pulses . Extracts were clarified by centrifugation . A 50 μl aliquot of each extract was set aside as ChIP input material . For the IP , 450 μl of lysis buffer was added to 50 μl of extract , and 5 μl of α-Wor1Nterm antibody was added . α-Wor1Nterm is an affinity-purified antibody generated against a peptide , QVLDKQLEPVSRRPHERER , located near the N terminus of Wor1 ( Bethyl Laboratories , http://www . bethyl . com/ ) . The IP was incubated for 2 h at 4 °C , with agitation . Then 50 μl of a 50% suspension of protein A-Sepharose Fast-Flow beads ( Sigma , http://www . sigmaaldrich . com/ ) in lysis buffer was added to the IP and incubated 1 . 5 h at 4 °C with agitation . The beads were pelleted for 1 min at 3 , 000 g . After removal of the supernatant , the beads were washed with a series of buffers for 5 min for each wash: twice in lysis buffer , twice in high-salt lysis buffer ( 50 mM HEPES-KOH [pH 7 . 5] , 500 mM NaCl , 1 mM EDTA , 1% Triton X-100 , 0 . 1% sodium deoxycholate ) , twice in wash buffer ( 10 mM Tris-HCl [pH 8 . 0] , 250 mM LiCl , 0 . 5% NP-40 , 0 . 5% sodium deoxycholate , 1mM EDTA ) , and once in TE ( 10 mM Tris , 1 mM EDTA [pH 8 . 0] ) . After the last wash , 100 μl of elution buffer ( 50 mM Tris-HCl [pH 8 . 0] , 10 mM EDTA , 1% SDS ) was added to each sample , and the beads were incubated at 65 °C for 15 min . The beads were spun for 1 min at 10 , 000 g , and the supernatant was removed and retained . A second elution was carried out with 150 μl of elution buffer 2 ( TE , 0 . 67% SDS ) and eluates from the two elution steps were combined . For the ChIP input material set aside , SDS ( 1% final concentration ) and 200 μl of TE were added . ChIP and input samples were incubated overnight at 65 °C to reverse the formaldehyde crosslinks . 250 μl of proteinase K solution ( TE , 20 μg/ml glycogen , 400 μg/ml proteinase K ) was added to each sample , and samples were incubated at 37 °C for 2 h . Samples were extracted once with 450 μl Tris buffer-saturated phenol/chloroform/isoamyl alcohol solution ( 25:24:1 ) . 55 μl of 4 M LiCl and 1 ml of 100% ethanol ( 4 °C ) were added and the DNA was precipitated for 1 h at 4 °C . The DNA was pelleted by centrifugation at 14 , 000 g for 15 min at 4 °C , washed once with cold 75% ethanol , and allowed to air dry . The samples were resuspended in 25 μl of TE containing 100 μg/ml RNaseA and incubated 1 h at 37 °C . ChIPs were also carried out in experiments not shown using affinity-purified antibody generated against a peptide DDAVGNSSGSYYTGT , located at the C terminus of Wor1 ( α-Wor1Cterm ) ( Bethyl Laboratories ) [11] . ChIP was performed in WT opaque strains twice using α-Wor1Nterm and three times using the α-Wor1Cterm antibodies . Control ChIPs were performed in the wor1Δ/wor1Δ mutants using α-Wor1Nterm once , and the α-Wor1Cterm was used twice . ChIP-enriched DNA was amplified and fluorescence labeled as described [32] . Labeled DNA for each channel was combined and hybridized to arrays in Agilent hybridization chambers for 40 h at 65 °C , according to protocols supplied by Agilent ( Agilent Technologies , http://www . agilent . com/ ) . Arrays were then washed and scanned , using an Axon Instruments Genepix 4000A scanner . Approximately 185 , 000 60-mer oligo probes were designed across the entire Candida genome ( contig20 haploid genome assembly ) at approximately 80 bp intervals , excluding nonunique regions of the genome ( see Protocol S1 for further information ) . Custom microarrays were manufactured by Agilent Technologies ( Agilent Technologies ) . Array design and ChIP-chip data are available on GEO . Arrays were blank subtraction normalized , interarray median normalized , and intra-array median normalized using Agilent ChIP Analytics 1 . 3 software ( Agilent Technologies ) . After normalization , average ratios across replicate arrays ( where relevant ) were used for further analysis . After normalization , the single array error model was applied across replicate arrays ( where relevant ) , to derive a p-value statistic to represent the probabilities that data at each spot occurred within experimental noise . A segment is a region of adjacent probes containing peaks of Wor1 enrichment , where the enrichment above input is considered to be statistically significant , based on the parameters set in the software . Using the ChIP Analytics software , the Whitehead Neighborhood Model was applied using default parameters as described [21] to map the segments according to their chromosomal positions . When comparing ChIP-chip experiments in WT opaque strains against wor1Δ/wor1Δ strains , or between α-Wor1 ChIP-chip experiments performed in WT opaque strains using the two different Wor1 antibodies , any overlapping segments were eliminated from further analysis . Within each segment , we used ChIP Analytics software to identify the location of highest Wor1 enrichment ( corresponding to the probe with the lowest P[ξ]-value ) . The positions of peaks were then assessed in relationship to ORFs throughout the C . albicans genome; an ORF was identified as being potentially regulated by Wor1 if there was a segment of Wor1 enrichment within the intergenic region immediately upstream of the given coding sequence .
Candida Genome Database ( http://www . candidagenome . org/ ) accession numbers for the genes discussed in this article are CZF1 ( orf19 . 3127 ) , WOR2 ( orf19 . 5992 ) , HAP3 ( orf19 . 4647 ) , orf19 . 4972 , CSR1 ( orf19 . 3794 ) , and PHO23 ( orf19 . 1759 ) .
|
The opportunistic fungal pathogen Candida albicans can switch between two heritable states—the “white” and “opaque” states . These two cell types differ in many characteristics , including cell structure , mating competence , and virulence . Recent studies of the molecular mechanism of regulating the white-opaque switch identified a master transcriptional regulator , Wor1 . In this study , we identified two transcriptional regulators , Czf1 and Wor2 , as new regulators of white-opaque switching . By constructing a series of single and double mutants and by examining where the master regulator Wor1 binds throughout the genome , we generated a molecular model of the bistable switch that regulates white-opaque switching . The regulatory model consists of interlocking positive feedback loops , which mutually reinforce one another and stabilize the opaque state . These results show how an organism can exist in two distinctive , heritable states without changes in the nucleotide sequence of its genome .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] |
[
"microbiology",
"molecular",
"biology",
"genetics",
"and",
"genomics"
] |
2007
|
Interlocking Transcriptional Feedback Loops Control White-Opaque Switching in Candida albicans
|
The rise of large-scale recordings of neuronal activity has fueled the hope to gain new insights into the collective activity of neural ensembles . How can one link the statistics of neural population activity to underlying principles and theories ? One attempt to interpret such data builds upon analogies to the behaviour of collective systems in statistical physics . Divergence of the specific heat—a measure of population statistics derived from thermodynamics—has been used to suggest that neural populations are optimized to operate at a “critical point” . However , these findings have been challenged by theoretical studies which have shown that common inputs can lead to diverging specific heat . Here , we connect “signatures of criticality” , and in particular the divergence of specific heat , back to statistics of neural population activity commonly studied in neural coding: firing rates and pairwise correlations . We show that the specific heat diverges whenever the average correlation strength does not depend on population size . This is necessarily true when data with correlations is randomly subsampled during the analysis process , irrespective of the detailed structure or origin of correlations . We also show how the characteristic shape of specific heat capacity curves depends on firing rates and correlations , using both analytically tractable models and numerical simulations of a canonical feed-forward population model . To analyze these simulations , we develop efficient methods for characterizing large-scale neural population activity with maximum entropy models . We find that , consistent with experimental findings , increases in firing rates and correlation directly lead to more pronounced signatures . Thus , previous reports of thermodynamical criticality in neural populations based on the analysis of specific heat can be explained by average firing rates and correlations , and are not indicative of an optimized coding strategy . We conclude that a reliable interpretation of statistical tests for theories of neural coding is possible only in reference to relevant ground-truth models .
Recent advances in neural recording technology [1 , 2] and computational tools for describing neural population activity [3] make it possible to empirically examine the statistics of large neural populations and search for principles underlying their collective dynamics [4] . One hypothesis that has emerged from this approach is the idea that neural populations might be poised at a thermodynamic critical point [5 , 6 , 7] , and that this might have consequences for how neural populations process sensory information [7 , 8] . As similar observations have been made in other biological systems [9 , 10 , 11] , it has been suggested that this might reflect a more general organising principle [12] . Critical phenomena play a central role in physics: Phase transitions mark a special point in which media qualitatively change their properties by transitioning from one state of matter into another ( e . g . liquid to gaseous at boiling point , ferro-magnetic and paramagnetic phases , or the emergence of super-conductivity ) . As such , the behaviour of a system at critical points is informative about its intrinsic properties . Moreover , critical points are ‘special’ in the sense that they classically only occupy a small portion of the parameter space . Thus , observing that a system is constantly poised at a critical point would be surprising , and would hint at an underlying organizing mechanism that keeps the system at this point . Given the fundamental importance of critical phenomena in physics , and their success in revealing the laws the determine the behaviour of physical systems , the hypothesis that these approaches might also shed lights on principles underlying neural coding is intriguing . Evidence in favour of this hypothesis has been put forward by a series of studies which measured neural activity from large populations of retinal ganglion cells and reported that their statistics resemble those of physical systems at a critical point [7 , 8] . To this end , Tkačik and colleagues developed a data analysis framework to search for signatures of criticality in experimentally obtained measurements . Using large-scale multielectrode array recordings [2] and maximum entropy models [13 , 14 , 15 , 16 , 17 , 12 , 3 , 18] , it was observed that the normalized variance of log-probabilities diverges as a function of population size . Importantly , this quantity is mathematically equivalent to the specific heat capacity , an important characteristic which diverges at critical points . In addition , when an artificial ‘temperature’ parameter was introduced , specific heat appeared to be maximal for the statistics of the observed data , rather than for statistics which have been perturbed by changing the temperature parameter . These properties of retinal populations resemble the behaviour of physical systems at critical points . It has been hypothesised [12 , 7] that the system needs to be optimized to keep itself at a critical point , for example through adaptation to stimulus statistics [19 , 20 , 21] or alternative mechanisms of self-organization [22 , 23 , 24] . A competing hypothesis states that instead generic mechanisms are sufficient to give rise to activity data with divergent specific heat , and that the presence of signatures of criticality does not provide evidence for retinal circuits being poised at a special state that is advantageous for coding . A series of theoretical studies [25 , 26 , 27 , 28] has shown that common input ( i . e , . the presence of latent variables ) can account for signatures of criticality: In particular , Schwab et al [26] and Aitchison et al [27 , 28] showed that Zipf scaling ( an alternative characterization of criticality ) and the divergence of the specific heat are closely related , and that in high-dimensional models with a low-dimensional latent variable , the specific heat diverges with system size under a wide range of circumstances [27 , 28] . Similarly , it has been shown empirically that a purely feedforward model can capture Zipf-like scaling in recordings from the salamander retina [29] . Interpreting findings of thermodynamic criticality for neural populations , identifying their mechanistic underpinnings , and clarifying their relationship with alternative theories , has been fraught with difficulty . We hypothesize that this difficulty stems from a subtle but crucial difference between how the scaling behaviour of system properties is studied in thermodynamics and in practical neural data analysis: Most theoretical approaches study how system properties scale as the size of the system , n , is varied . In contrast , in practical neural data analysis , different “n” do not correspond to different system sizes , but are obtained by subsampling neural populations from a large recording ( which is itself a subsample of the underlying system ) . How does this sampling process affect estimates of whether the system is at a critical point ? A second difficulty in interpreting these studies stems from the fact that they are based global statistical measures whose relationship with simple statistics such as firing rates and correlations— which are commonly used and have been extensively studied in neural coding [30 , 31]—is unclear . We here focus on one statistic that has been used as evidence of critical behaviour , namely the dependence of specific heat on population size and temperature . We study how it depends on neural firing rates and correlations , as well as on how this data is subsampled during data analysis: First , we show explicitly that signatures of criticality , can be reproduced in canonical feed-forward models of neural population activity , as predicted by previous studies [25 , 26 , 28] . These studies did not have tools for studying population statistics in large simulations , and they were therefore limited to studying small ( n ≤ 40 ) systems– for these small system sizes , it is difficult to make statements about the peak in the specific heat and its scaling with population size . In particular , the dominant peak near unit temperature only emerges for much larger systems . We overcome this difficulty by providing improved algorithms for efficiently fitting maximum entropy models to large neural populations ( available at https://github . com/mackelab/CorBinian ) , and use them to apply the analyses proposed by previous studies [7] to data simulated from a simple , feedforward encoding model of retinal processing [32 , 33 , 34 , 35] . Second , previous theoretical studies [26 , 27 , 28] treated only the limiting behavior of the specific heat at unit temperature , and did not investigate its dependence on firing rates and correlations . We here relate the characteristic shape of specific heat curves ( i . e . the dependence of specific heat on temperature ) to neural correlations and firing rates . The emergence of peak specific heat at the ‘inherent’ temperature T = 1 has given rise to the idea that correlations in the observed system are ‘special’ , i . e . that systems with stronger or weaker correlations would not exhibit them [7] . We use an analytically tractable model of the analysis process to show that this is not the case– the more strongly correlated the population is , the more pronounced signatures of criticality will be . This analysis also shows that a ‘low-temperature’ regime ( as reported by [18] ) will be found whenever firing rates are sufficiently low . Third , we analyze the structure of correlations which are sufficient to induce signatures of criticality , and find that it is sufficient if the average correlation is independent of population size . Such ‘criticality-inducing’ correlations can arise both from neural mechanisms such as common input or dense connectivity . Importantly , we show that they can also arise as a consequence of data analysis: Uniformly subsampling a recording with any non-zero correlations to construct subpopulations yields criticality-inducing correlations . In summary , we show that statements about signatures of criticality derived from thermodynamics can be reduced to statements about firing rates and correlations , and that correlation structures which give rise to these signatures are ubiquitous in neural populations .
A hallmark of criticality is that the specific heat capacity of the model diverges when the temperature reaches the critical temperature [5] . Tkačik et al . [7] developed an approach for translating this concept to neural data analysis ( see Fig 1 ) : . In this analysis , neural populations of different size n are generated from the full recording ( of size N ) by random subsampling . The statistics of activity for each population of size n are characterized using a maximum entropy model fit to population activity [13 , 14 , 16 , 17 , 3] . Finally , the maximum entropy models are perturbed by introducing a temperature parameter , and specific heat is computed for each population size n and temperature T from the ( perturbed ) maximum entropy model fit . Divergence of specific heat with population size n , and a peak of the specific heat near unit temperature T = 1 ( the ‘temperature’ of the original data ) are interpreted as indication for the system being at a critical point [7] . We wanted to verify that this phenomenon could be captured in feedforward models of retinal processing . We wanted to directly demonstrate that canonical mechanisms of retinal processing—such overlapping centre-surround receptive fields , spiking nonlinearities , shared Gaussian noise—are sufficient for the signatures of criticality to arise . We first created a simple phenomenological model of retinal ganglion cell ( RGC ) activity based on linear-nonlinear neurons [32 , 33 , 35] . In this model ( Fig 2a ) , we assumed retinal ganglion cells to have centre-surround receptive fields [36 , 35] with linear spatial integration [37] , sigmoid nonlinearities and stochastic binary spikes: in each time bin of size 20ms , each neuron i either emitted a spike ( xi = 1 ) or not ( xi = 0 ) . We used a sequence of natural images as stimuli . In addition to the feedforward drive by the stimulus , nearby neurons received shared Gaussian noise , mimicking common input from bipolar cells [30] . Thus , cross-neural correlations in the model arise from correlations in the stimulus , receptive-field overlap and shared noise , but not from lateral connections between RGCs . As we will explain below , only the strength of correlations , but not their mechanistic origin or dependence on stimuli , is relevant for determining the specific heat . Parameters of the model were chosen to approximate the statistics of receptive-field centre locations of RGCs , as well as histograms of firing rates , pairwise correlation-coefficients and population spike-counts ( Fig 2b ) . We subsampled populations of different sizes 20 ≤ n ≤ 120 by uniformly sampling cells from our simulated recording of total size N = 316 neurons . For each population we fit a ‘K-pairwise’ maximum entropy model [3] . This model assigns a probability P ( x ) to each spike-pattern x . It is an extension of pairwise maximum entropy models ( i . e . Ising models ) [13 , 14] which reproduce the firing rates and pairwise covariances , and has additional terms to capture population spike-counts [3] ( see Materials for details of model specification and parameterisation ) . As we needed to efficiently fit this model [38 , 39] to multiple simulated data sets , we developed an improved fitting algorithm ( see section 1 in S1 Supporting Information ) based on maximum-likelihood techniques using Markov chain Monte Carlo ( MCMC ) , building on work by [15] . In particular , we made the most computationally expensive component of the algorithm , the estimation of pairwise covariances via MCMC sampling , more efficient by using a ‘pairwise’ Gibbs-sampling scheme with Rao-Blackwellisation [40] ( see section 1 . 1 in S1 Supporting Information ) . Most Gibbs-sampling approaches for maximum entropy models [15] update one neuron i at a time by re-sampling its state from the conditional distribution , given the state of the other n − 1 neurons in the population . We here in each iteration update a randomly chosen pair ( i , j ) simultaneously , given the state of the other n − 2 neurons . While each pairwise sample is more expensive to compute , this approach has the advantage of yielding a direct estimate of the ( conditional ) probability of i and j being active simultaneously . From these conditional probabilities , one can estimate pairwise covariances more efficiently than is possible through averaging samples , a process which is known as Rao-Blackwellization . Here , Rao-Blackwellization resulted in a reduction of the number of samples ( and computation time ) needed for achieving low-variance estimates of the covariances by a factor of approximately 3 ( Fig 2c , Fig . A in S1 Supporting Information ) . After parameter fitting , the model reproduced the statistics of the simulated data ( Fig 2d , Fig . B in S1 Supporting Information ) . Following [7] , we then introduced a temperature parameter which rescales the probabilities of the model , P T ( x ) ∝ P ( x ) 1 / T , ( 1 ) where temperature T = 1 corresponds to the statistics of the empirical data . By changing T to other parameter values one can perturb the statistics of the system [41]: Increasing temperature leads to models with higher firing rates and weaker correlations ( Fig . C in S1 Supporting Information ) , with PT ( x ) approaching the uniform distribution for large T . If the temperature is decreased towards zero , PT ( x ) has most of its probability mass over the most probable spike patterns . We compute the specific heat of a population directly from the probabilistic model fit to data [7] , using c ( T ) = 1 n Var [ log P T ( X | λ ) ] , ( 2 ) i . e . the variance of the log-probabilities of the model with parameters λ , normalised by n . While specific heat is typically motivated by thermodynamics , in this context it corresponds to a global statistical measure which provides a compact mathematical description of the collective statistical dynamics of the system . Just like the entropy corresponds to the ( negative ) average log-probability across all population states , the specific heat corresponds to the ( normalized ) variance of log-probabilities . Thus , specific heat is minimal for data in which all patterns x are equally probable , and big for data in which pattern-probabilities span a large range . We used MCMC-sampling to approximate the variance across all probabilities , and used this approach to calculate , for each population of size n , the specific heat as a function of temperature ( Fig . D in S1 Supporting Information ) . We found that the temperature curves obtained from the simulated data qualitatively reproduce the critical features of those that had been observed for large-scale recordings in the salamander [7] and rat [8] retina: The peak of the curves diverges as the population size n is increased , and moves closer to unit temperature for increasing n ( Fig 2e ) . Consistent with experimental findings [42 , 7 , 8] ( Fig 2f ) and [28] , we found that specific heat diverged linearly with population size . Finally , and also consistent with experimental studies , the peak specific heat is achieved for T > 1 , which is what has been interpreted as a ‘low-temperature’ state [18] . These results confirm that signatures of criticality arise in a simple feedforward LN cascade model based on generic properties of retinal ganglion cells , and do not require finely tuned parameters or sophisticated circuitry . In the phenomenological population model above , we observed that specific heat grew linearly with population size , as it did in previous studies built on experimental data [42 , 7 , 8 , 18] . Different ‘populations’ in these analyses are obtained by subsampling different populations from a large experimental recording , and that the parameters of each of these models are independently fit to each such population . How does this analysis process effect the rate of divergences of the specific heat , and the qualitative shape of specific heat curves ? To answer these questions , we build a simple mathematical description of the analysis process: In the original papers , populations of different sizes are obtained by randomly subsampling a large recording ( which is itself a sub-sample of the underlying circuit ) . As the simplest possible description of this sampling process , we assume that there is an underlying , infinitely large neural population , and that each population of size n is a random subsample . We assume that the underlying population is homogeneous , i . e . that all neurons have the same mean firing rate and pairwise correlations . As a consequence , K-pairwise maximum entropy models are fully specified by the distribution of population spike-count K = ∑i xi [25 , 43 , 44 , 45] for each population of size n . We refer to models with this property as ‘flat models’ ( [46] calls them ‘reduced’ maximum entropy models ) . We introduce a new parametrised flat model in which the spike-count distribution is given by the beta-binomial distribution P ( K|α , β , n ) , reducing the number of free parameters from n to 2 . The beta-binomial model is a straightforward extension of an independent ( i . e . binomial ) population model: At each time-point , a new firing probability p is drawn from a beta-distribution with parameters α and β , and neurons then spike independently with probability p . Fluctuations in the latent variable p are shared across the population and lead to correlations in neural activity . Therefore , this model is a particular instance of a latent variable model . Signatures of criticality in latent variable models have been studied previously [26 , 27 , 28] . Our analytically-tractable model provides an explicit construction of how subsampling a large population determines the dependence of specific heat on population size . Our beta-binomial model provided a good fit to the population spike-count distributions of the simulated data ( Fig 3a ) across different population sizes n ( Fig 3b ) . Importantly , the best-fitting parameters α and β did not vary systematically across population sizes , and converged to values of α = 0 . 38 and β = 12 . 35 ( Fig . E in S1 Supporting Informationa ) , corresponding to a probability of spiking of μ = 0 . 03 in each bin ( i . e . each neuron has an average firing rate of μ/Δ = 1 . 5 Hz ) and average pairwise correlations of ρ = 0 . 073 . The beta-binomial model also provided good fits to published population spike-count distributions [43 , 45 , 8] , as well as to those of retinal ganglion cell activity under different stimulus conditions in [18] ( Fig . E in S1 Supporting Information ) . When we applied this flat model to populations subsampled from the RGC simulation , we could qualitatively reproduce the specific heat curves of the K-pairwise model ( see also Fig . F in S1 Supporting Information ) . In particular , we found a linearly diverging peak that moved closer to T = 1 as the population size was increased ( Fig 3c ) . Thus , linear divergence of specific heat is qualitatively captured by this model of how different populations are obtained by subsampling a large population . One of the difficulties of interpreting the scaling behaviour of maximum entropy models fit to neural data is the fact that the construction of the limit in n differs from those studied in statistical physics: In statistical physics , different ‘n’ typically correspond to systems of different total size , and the parameters are scaled as a deterministic function of n ( e . g . drawn from a Gaussian with variance proportional to 1/n in spin-glasses [47 , 48] ) . In studies using maximum entropy models for neural data analysis , populations of different n are obtained by randomly subsampling a fixed large recording , and the parameters are fit to each subpopulation individually . Thus , there is no analytical relationship between population size and parameter values in this approach . With our model of the analysis process based on flat models , it is possible to analytically characterise the behaviour of the specific heat for large population sizes for this sampling process [25 , 44] . Using this approach , one can show ( section 2 . 3 in S1 Supporting Information and [25] for details ) that for virtually all flat models , the specific heat diverges linearly at unit temperature , but not for any other temperature T > 1 or T < 1 ( section 2 . 4 in S1 Supporting Information ) . As a consequence , the peak must move to T = 1 as n is increased . Hence , almost any flat model analysed with the methods developed by [7] will exhibit signatures of criticality . In particular , these results hold also for models which are more weakly or more strongly correlated than real neural populations , and even for models with unrealistic population spike-count distributions ( see Fig . G in S1 Supporting Information for an illustration ) . There are only two exceptions: The first one is a model in which all neurons are independent ( i . e . a binomial population model ) , and the second one is a flat pairwise maximum entropy model—indeed , this is the only flat model with non-vanishing correlations for which the specific heat does not have its peak at unit temperature ( see [25] for an illustration for the flat pairwise maximum entropy model ) . Finally , it has been observed that the peak of the specific heat curve is consistently ‘to the right’ of T = 1 , which was interpreted as the neural population activity in the retina being in a ‘low-temperature state’ [18] . Our analysis based on the flat model gives insights into this phenomenon: For correlation ρ = 0 , the position of the peak can be calculated in closed form ( Fig 3d ) . We observe that the peak will be at temperatures >1 whenever the spike probability is smaller than μ* = 0 . 0832 , which corresponds to a firing rate of μ*/Δ = 4 . 16Hz at a bin size of Δ = 20ms . Thus , in our model , the ‘temperature-state’ of a population can be reduced to a statement about the firing rate relative to the bin size used for analysis: For ρ > 0 ( Fig 3e ) and for larger population sizes n , the firing rate at which the transition occurs are shifted to slightly higher firing rates , i . e . the ‘low-temperature’ regime is even bigger , and e . g . extends to firing rates up to 8 . 63Hz for average correlations of ρ = 0 . 25 and population size n = 120 ( Fig 3f ) . While this dependence may be more complicated for full correlation structures , our analysis again connects global population measures from statistical mechanics to basic , directly measurable statistics of neural data: ‘being in a low-temperature state’ is a statement about the firing rates in the population being low . The rate at which the specific heat diverges provides a mean of quantifying the ‘strength’ of criticality . What is the relationship between correlations in a neural population and the rate of divergence ? To study how the specific heat rate c ˜ = c ( T = 1 ) / n depends on the strength of correlations , we used a beta-binomial model to generate simulated data with firing rate μ/Δ = 1 . 5Hz ( i . e . each neuron has a probability of spiking of μ = 0 . 03 per bin ) , and different pairwise correlation coefficient ρ ranging from ρ = 0 . 01 to ρ = 0 . 25 ( Fig 4a ) . The heat curves had the same shape as in the analyses above , with a peak that increases and moves to unit temperature ( Fig 4b ) . We found that the specific heat rates c ˜ increased strictly monotonically with ρ ( Fig 4b and 4c ) . For the beta-binomial model , the large-n value of c ˜ can be calculated analytically ( section 3 . 2 in S1 Supporting Information for details ) as a function of the parameters α and β , c˜=α ( α+1 ) ψ1 ( α+1 ) +β ( β+1 ) ψ1 ( β+1 ) ( α+β ) ( α+β+1 ) +αβ ( ψ0 ( α+1 ) −ψ0 ( β+1 ) ) 2 ( α+β ) 2 ( α+β+1 ) −ψ1 ( α+β+1 ) , ( 3 ) where ψ0 , ψ1 denote the di- and trigamma function , respectively . This analytical evaluation of c ˜ ( valid for large n ) was in good agreement with numerical simulations ( Fig 4c left ) . In the case of weak correlations ρ , eq 3 can be simplified: In this case , the specific heat rate is proportional to the strength of correlations ( section 3 . 1 in S1 Supporting Information for details ) , i . e . c˜≈ρμ ( 1−μ ) ( log ( 1−μμ ) ) 2 , ( 4 ) and also increases strongly with firing rate for small μ ( Fig . H in S1 Supporting Information ) . This expression can also be derived from the Gaussian model in [8] equation ( 4 ) , by inserting the expected values of the mean and variance of the population spike-count under random subsampling . The monotonic relationship between correlations and specific heat is also consistent with the derivation in [27] for latent-variable models: inspection of equation ( 65 ) in [27] shows that the specific heat is related to a sum of conditional entropies– for binary random variables , these entropies are monotonically related to covariances , which effectively shows that , in their model , specific heat also increases with correlations . We found that the relationship between the strength of correlations and the ‘strength’ of criticality ( i . e . the divergence rate of specific heat ) also held in simulations of feedforward models of retinal population activity . In the original study [7] , specific heat was computed from K-pairwise model fits to RGC activity resulting from three different kind of stimuli: random checkerboard stimuli ( which do not have long-range spatial correlations , although stimulus-driven cross-neural correlations can arise from receptive field overlap ) , natural stimuli , which exhibit strong spatial correlations , and full-field flicker ( which constitutes an extreme case of spatial correlations since all pixels in the display are identical ) . It was found that specific heat diverges in all three conditions ( consistent with a more recent study [18] ) , and interpreted this as evidence that signatures of criticality are not ‘inherited from the stimulus’ [7] . When we simulated responses to different stimuli we found the divergence rates of the specific heat to follow the pattern of induced correlation strength , consistent with the monotonic relationship between correlation strength and specific heat growth rate shown above for the flat models ( Fig 4d ) : For populations size n = 100 , checkerboard/natural/full-field flicker stimulation lead to average correlation strengths of ρ = 0 . 033/0 . 075/0 . 341 , respectively , and to specific heat growth rates of c ˜ = 0 . 0029/0 . 0046/0 . 0104 . Tkačik et al . had found the lowest peak in divergence rate for checkerboard ( max c ≈ 0 . 54 ) , higher peak-divergence rates for natural movies ( max c ≈ 0 . 92 ) and the highest peak for full-field flicker ( max c ≈ 2 . 4 , all results for n = 100 ) . Thus , the ordering of the peak values of specific heat in their study is consistent with our results . However , when comparing the values at T = 1 , they found a slightly higher divergence rate for natural movies ( c ˜ ≈ 0 . 005 ) than for full-field flicker ( c ˜ ≈ 0 . 004 ) . This mismatch could result from adaptation or temporal dynamics of the stimulus affecting firing rates or correlations in their data [20] , or from our simulations not precisely matching the statistics of their experimental data . These statements also qualitatively hold in a modified temperature analysis [7] in which firing rates are kept constant ( at the firing rates of T = 1 ) when temperature is varied ( section 3 . 4 in S1 Supporting Information and in Fig . I in S1 Supporting Information ) . We conclude that the experimental evidence—which showed that the specific heat diverges , and how the speed of divergences depends on the stimulus ensemble—is largely consistent with a simple , feedforward phenomenological model of retinal processing . Thus , at least for flat models , ‘being very critical’ is a consequence of ‘being strongly correlated’ , and not evidence for correlations being fine-tuned or self-organized to a particular value . In the above , we showed that a beta-binomial spike-count distribution can be sufficient for signatures of criticality to arise . For this to hold we need the variance of the population spike-count to grow quadratically with population size , i . e . Var ( K ) ∝ n2 . The variance of the population spike-count is equal to the sum of all variances and covariances in the population , Var ( K ) = ∑ i = 1 n Var ( x i ) + ∑ i ≠ j Cov ( x i , x j ) . A sufficient condition for signatures of criticality to arise in these models is that the average covariances ( and hence correlations ) between neurons are independent of n , 1 n ( n - 1 ) ∑ i ≠ j Cov ( x i , x j ) ≈ constant [27 , 6 , 5] . We refer to correlations with this property as ‘criticality inducing’ . One possible criticality-inducing correlation structure are so called ‘infinite range’ correlations: correlation between neurons do not drop off to zero for large spatial distances . In the extreme case of distance-independent correlations ( Fig 5a ) , adding more and more neurons to a population will not change the average pairwise correlation within the population ( Fig 5b ) . We note that infinite-range correlations are typically not present in the thermodynamic limit in physical systems at equilibrium . In neural systems , infinite-range correlations could be a consequence of densely connected circuitry , or of a shared stimulus drive . Importantly , criticality-inducing correlations can also result as a consequence of subsampling a large neural population: Even a neural population which does not have infinite-range correlations can appear critical if it is randomly subsampled during analysis . If different populations of size n are obtained as above by ( uniformly ) subsampling a large recording of size N , then the pairwise correlations in each subpopulation are also a random subsample of the large correlation matrix of the full recording . For any correlation structure on the full recording ( including limited-range correlations , Fig 5c ) , the expected average correlation in a population of size n is identical to the average correlation in the full recording and hence independent of n ( Fig 5d left , grey line ) . Despite the pairwise correlations being subsampled in blocks of principal submatrices rather than independently , the variance of the average correlation can drop with the square of the population size n , and is guaranteed to fall at least as 1/n ( section 4 . 1 in S1 Supporting Information , and Fig . J in S1 Supporting Information ) . Because the average correlation will be independent of n and have negligible variance ( Fig 5d left , shaded area ) , specific heat will diverge with constant slope ( Fig 5d right ) . In contrast , if different population sizes are constructed by taking into account the spatial structure of the population ( i . e . by iteratively adding neighbouring cells ) then the average correlation in each subpopulation will drop with n , and the slope of specific heat growth will decrease with population size . In our RGC simulation , pairwise correlations did drop off to zero with spatial distance for checkerboard and natural images , but not for full-field flicker ( Fig 5e ) . Pairwise correlations in the full-field flicker condition initially drop off due to distance-dependent shared noise , but eventually saturate at a level far above zero that is determined by the full-field stimulus . Due to these strong infinite-range correlations , both spatially structured sampling and uniform sampling then give rise to linear growth in specific heat ( Fig 5f left ) . For the other two stimulus conditions , however , the choice of subsampling scheme does result in markedly different behavior of the specific heat growth: Both for natural images and checkerboard stimuli , we can see the rate of growth decreases for large n under spatially structured subsampling ( Fig 5f centre and right ) . This effect will be more pronounced for larger simulations , and in additional simulations we found specific heat to saturate once populations are substantially bigger than the spatial range of correlations . This behavior is not unique to the simplified flat models . Specific heat traces computed from K-pairwise models fit to populations obtained with spatially structured sampling also show a marked decrease in specific heat growth rates ( section 4 . 2 in S1 Supporting Information and Fig . K in S1 Supporting Information ) . In summary , populations will exhibit critical behaviour if correlations have infinite range ( over the size of the recording ) , irrespective of the sampling scheme . In addition , if a population is randomly subsampled ( as was done in [7 , 8] ) , then signatures of criticality will arise even if the underlying correlations have limited range .
An intriguing hypothesis about the collective activity of large neural populations has been the idea that their statistics resemble those of physical systems at a critical point . In recent years , several studies [12 , 5 , 6 , 11 , 7 , 8 , 18] proposed a new approach to studying criticality in biological data , motivated by notions of criticality in thermodynamics . Signatures of criticality have also been observed in natural images [11] and cortical populations [6] , and have been studied using the theory of finite-size scaling and critical exponents [6] . It has been argued that systems close to a critical point might be optimally sensitive to external perturbations [6] and that the large dynamic range of the code ( i . e . large variance of log-probabilities ) might be beneficial for encoding sensory events which likewise have a large distribution of occurrence probabilities [16] . This hypothesis that neural systems are poised at a thermodynamic critical point could open up further questions on how the system maintains its critical state and on implications for how neural populations encode sensory information and perform computations on it . Alternatively , generic mechanisms could be sufficient to give rise to data which satisfies the definition of criticality put forward in these studies . We had demonstrated in a previous theoretical study [25] that simple models with Gaussian common input can exhibit a diverging specific heat . More recently , it was shown [26 , 27 , 28] that common input ( or other latent variables which lead to shared modulations in firing rates , such as non-stationarity [29] ) can give rise to Zipf-like scaling of pattern probabilities , a second signature of criticality . Mathematically , Zipf’s Law is equivalent to stating that the plot of entropy vs energy ( i . e . log-probability ) is a straight line with unit slope [26 , 27] . Schwab et al [26] showed that particular latent variable models give rise to Zipf’s law . This result was generalized [27 , 28] to show that , under fairly general circumstances , high-dimensional latent variable models exhibit a wide distribution of energies ( i . e . log-probabilities ) and hence a large specific heat . It has also been argued that the use of data sets which are too small might give rise to spuriously big specific heats [49]: while this could be true in principle , additional analyses e . g . in [7] show that their results are robust with respect to data set size , and our results are also valid even in the case of infinite data . Finally , it has also been suggested that whether statistical models exhibit criticality depends on which variables are measured and constrained by the model fit [50 , 51] . Previously , criticality in neural systems has also been investigated extensively using a definition of criticality which is based on temporal dynamics with power-law statistics , so-called ‘avalanches’ [52 , 5] . Numerous studies have reported and studied ‘avalanche criticality’ [8 , 21] , proposed possible mechanisms ( e . g . based on self-organization [53] ) , and discussed finite-size effects and sub-sampling [54] , as well as a need for rigorous statistical analysis [55] . We emphasize that the ‘avalanche’ definition of criticality is not equivalent to the thermodynamics-inspired definition used in these more recent studies [12 , 8] . Our study is only concerned with this more recent approach , and our results thus have no bearing on studies of ‘avalanche-criticality’ . We here related signatures of criticality to the structure of firing rates and correlations in the population: We found that average correlations which are independent of population size are sufficient for inducing criticality , irrespective of their origin . In the thermodynamic analysis of physical systems at equilibrium , long-range correlations typically vanish in the thermodynamic limit . In neural systems , however , ‘criticality-inducing’ correlations can arise as a consequence of various factors: First , in a local patch of retina , retinal ganglion cells have a large degree of receptive field overlap , and natural stimuli also contain strong spatial correlations . This can lead to correlations which do have unlimited range within the experimentally accessible length scales . Thus , fluctuations in the stimulus will lead to common activity modulations amongst neurons within the population . Empirically , correlations between pairs of retinal ganglion cells only fall off slowly with the distance between somata ( or receptive field centres ) [35] . Second , firing rates e . g . of cortical neurons are modulated by global fluctuations in excitability [45 , 56] , resulting in neural correlations with infinite range . Third , and importantly , we showed that criticality-inducing correlations can also arise as a consequence of data analysis choices: Uniformly subsampling a large recording with correlations to construct subpopulations yields criticality-inducing correlations , even if the correlations itself do not have unlimited range . We also showed that there is a direct relationship between ‘how critical’ and ‘how correlated’ a population is: The stronger correlations are , the more prominent the divergence in specific heat is . Mechanisms underlying correlations in spiking activity have been extensively studied in neuroscience [30 , 31] , and our study makes it possible to relate ‘signatures of criticality’ derived from thermodynamics to these studies , and to interpret the significance of observing these effects: Given the ubiquity of criticality-inducing correlations , signatures of criticality are likely going to be found not just in retinal ganglion cells , but in multiple brain areas and model systems . They are entirely consistent with canonical properties of neural population activity , and require neither finely-tuned parameters in the population , nor sophisticated circuitry or active mechanisms for keeping the system at the critical point . The relationship between firing rates , correlations and criticality ( eqs 3 and 4 ) also yields a prediction about how adaptation in a classical sense should modulate signatures of criticality: The height of the peak is monotonically related to both correlation strength and firing rate . Adaptation typically reduces firing rates and correlations [57 , 58] . Taken together , this leads to the prediction that adaptation should reduce signatures of criticality– this is precisely the opposite of what has been predicted in [7] . Finally , the dependence of specific heat on correlations might also be an explanation of why Ioffe and Berry [18] found that a feedforward model fit to their retinal data ( which had lower correlations ) underestimated the specific heat . In summary , we conclude that current attempts to interpret findings of thermodynamic criticality in neural population activity have limited potential to lead to new insights into theories of neural computation– in particular , they are not able to discriminate between different hypotheses about either the origin or the functional consequence of the statistics of neural activity . A reliable interpretation of any test for criticality is possible only in reference to a-priori knowledge about the outcome of the test on relevant ground truth models . In order to realise the potential of large-scale recordings of neural activity in the search of a theory of neural computation , we will need data analysis methods which are adapted to the specific properties of biological data , and in particular the fact that neural activity is highly subsampled [59 , 60 , 54 , 61] . One approach to dealing with subsampled data is to use latent-variable models which explicitly model the effect of unobserved inputs and states [62 , 63] . In addition , we will also require hypotheses about the normative principles which govern their computations . A possible link between neural activity and theories of criticality might emerge from recent work in machine learning , which is starting to study links between the information-processing capabilities of artificial neural networks and critical phenomena [64] .
We simulated a population of N = 316 retinal ganglion cells as linear threshold neurons whose receptive fields were modelled by difference-of-Gaussian filters with ON-centres [37 , 35 , 33] . The simulation comprised two subgroups of cells with different receptive field sizes ( surrounds 56μm and 30μm in retinal space , centres 28μm and 15μm , respectively , one third cells with large receptive fields ) . For both subgroups , the weight of the surround was 0 . 5 of the centre weight . Locations of receptive field centres ( Fig 1 left panel ) were based on a reconstruction of 518 soma locations from a patch of mouse retina [65] . As the reconstructed locations in that data set also comprised about 40% amacrine cell somata , we randomly discarded 40% of the cell locations . The resulting patch of retina covered an area of 200 × 300μm2 , corresponding to 100 × 150 pixels in stimulus space . Correlated noise across neurons was modelled using correlated additive Gaussian noise . Correlations dropped off exponentially with soma distance with a decay constant of τ = 30μm i . e . noise covariance matrix was chosen as Σ = σ n o i s e 2 ( a I n + b e - D / τ ) , where Dij is the distance between neurons i and j and a2 + b2 = 1 . We set σnoise = 0 . 022 and a = 0 . 45 . We modelled neural spiking in discrete time using 20ms bins . In each bin t , the total input zi ( t ) to neuron i was given by z i ( t ) = w i ⊤ s ( t ) + ϵ i ( t ) , where wi is the receptive field of neuron i , s ( t ) the vectorised stimulus and ϵi ( t ) the input noise of neuron i . A neuron in a given bin is active ( xi = 1 ) if zi + d > 0 . 5 and inactive ( xi = 0 ) otherwise , with offset d = 0 . 168 [66] . Parameters of the simulation ( centre and surround sizes , relative strength of centre and surround , magnitude and correlations of noise , spiking threshold ) were chosen to roughly match the statistics of neural spiking ( firing rates , pairwise correlations , population activity counts ) reported in studies of salamander retinal ganglion cells [13 , 3 , 2] . We used three types of stimuli for this study: natural images , checkerboard patterns and full-field flicker . For natural image stimuli , we used a sequence of 101 images of foliages . Each image was 400 × 400 pixels , and each image was presented for 20ms with 300 repetitions total . The luminance histograms of the images were transformed to a normal distribution with mean 0 . 5 and pixel values between 0 and 1 . For the full-field flicker stimulus , luminance levels were drawn from a Gaussian distribution with mean μ = 0 . 5 and variance σ2 = 0 . 06 . Checkerboard stimuli consisted of 80 × 80 tiles of size 5 × 5 pixels each . Luminance levels ( from within the interval [0 , 1] ) of each tile were chosen to be either 0 . 15 or 0 . 77 with probability 0 . 5 . The parameters of both stimulus sets were chosen to match the dynamic range of the simulated retinal ganglion cells . For both types of stimuli , 2000 images were generated and the image sequences were presented with 10 repetitions . To calculate specific heat as function of increasing population size , we randomly selected 10 subsamples of the full simulated population of N = 316 cells at population sizes n ∈ {20 , 40 , 60 , 80 , 100 , 120} by uniformly drawing n neurons out of the full population without replacement . We modelled retinal ganglion cell activity by using a ‘K-pairwise’ maximum entropy model [3] . In a maximum entropy model [67] , the probability of observing the binary spike word x ∈ {0 , 1}n for parameters λ = {h , J , V} is given by P ( x | λ ) = 1 Z ( λ ) exp ( h ⊤ x + x ⊤ J x + ∑ k = 0 n V k δ ( K ( x ) = k ) ) ( 5 ) Here , the parameter vector h ( of size n × 1 ) and the upper-triangular matrix J ∈ R n × n correspond to the bias terms and interaction terms in a pairwise maximum entropy model ( also known as an Ising model or spin-glass ) [13] . The term K ( x ) = ∑ i = 1 n x i denotes the population spike-count , i . e . the total number of spikes across the population within a single time bin , and the indicator-term δ ( K = k ) is 1 whenever the population spike-count equals k , and is 0 otherwise . The term ∑ k = 0 n V k δ ( K = k ) was introduced [3] to ensure that the model precisely captures the population spike-count distribution of the data using n additional free parameters . The partition function Z ( λ ) is chosen such that the probabilities of the model sum to 1 . To fit the model parameters λ = {h , J , V} to a data set , we maximised the penalised log-likelihood [68 , 69] of the data D = { x ( 1 ) , x ( 2 ) , … , x ( M ) } under the model , L ( h , J , V ) :=∑m=1MlogP ( x ( m ) |h , J , V ) −1σh‖h‖1−1σJ‖J‖1−12VTΣ−1V . ( 6 ) Here , the l1-penalty controlled the magnitudes of parameters h , J , the term ‖J‖1 favoured sparse coupling matrices , and the regularisation term Σ on the V-parameters ensures that the terms controlling the spike-count distribution vary smoothly in k ( section 1 in S1 Supporting Information ) . This smoothness prior is particularly important for large spike counts , as it makes it possible to interpolate parameters for which the number of observed counts is small . In maximum entropy models , exact evaluation of the penalised log-likelihood and its gradients requires the calculation of expectations under the model , E[xi] , E[xi xj] or equivalently cov ( xi , xj ) , and P ( K = k ) ( section 1 . 1 in S1 Supporting Information ) , which in turn requires summations over all 2n possible states x and is prohibitive for n > 20 . Following previous work [15] , we used Gibbs sampling to approximate the relevant expectations ( section 1 . 1 in S1 Supporting Information for derivations and implementation details ) . We used two modifications over previous applications of Gibbs sampling to fitting maximum entropy models to neural population spike train data , with the goals of speeding up parameter learning and alleviating memory usage: First , we use Rao-Blackwellisation [40] to speed up convergence of the estimation of covariances of x: for this , we used pairwise Gibbs sampling ( blocked Gibbs with block size 2 ) , where each new sample in the MCMC chain was obtained by updating two entries i and j of x at a time , rather than just a single entry . This allowed us to get estimates of the conditional probabilities P ( xi xj = 1|x∼{i , j} ) , and to use them to speed up the estimation of the second moment E[xi xj] from empirical average of these conditional probabilities ( section 1 . 1 in S1 Supporting Information ) . Second , we used a variant of coordinate ascent that calculated all relevant quantities as running averages over the MCMC sample , and thereby avoided having to store the entire n × M ˜ MCMC sample in memory [15] , where M ˜ is the length of the sample . Because all features of the maximum entropy model are either 0 or 1 ( xi , xi xj and the indicator function for the spike count ) , the gain in log-likelihood obtainable from either updating a single element of h or J [15 , 39] , or from updating all V simultaneously ( but not from updating multiple entries of h and J ) can be computed directly from MCMC estimates of E[xi] , E[xi xj] and P ( K = k ) ( section 1 . 2 in S1 Supporting Information ) . For each iteration , we calculated the gain in log-likelihood for each possible update of hi , Jij and full V , and picked the update which led to the largest gain [15] . We measured the length of Markov chains in sweeps , where one sweep corresponds to one round of n ( n − 1 ) /2 Markov chain updates that encompasses all pairs of entries of x in random order . We set a learning schedule that started at 800 sweeps for the first parameter update and doubled the number of sweeps in the chain after each set of 1000 parameter updates . We monitored convergence of the algorithm using a normalised mean square error between empirical E[xi] , cov ( xi , xj ) , P ( K = k ) and their estimates from the MCMC sample . For normalisation , we used the average squared values of the target quantity , e . g . 1n∑i=1nE[ xi ]2 for the firing rates . We stopped the algorithm when a pre-set threshold was reached ( 0 . 01% , 0 . 25% , 0 . 01% for E[xi] , cov ( xi , xj ) , P ( K = k ) , respectively ) , or when the fitting algorithm took more than ( n100 ) 2×72h of computation time on a single core ( 2 . 294 GHz AMD Opteron ( TM ) Processor 6276 ) ( Fig . A in S1 Supporting Information ) . For 10 populations of size n = 100 ( for natural images ) , the normalised MSEs after model-fitting were 0 . 43% , 2 . 80% , 0 . 42% ) . An implementation of the fitting algorithms in MATLAB is available at https://github . com/mackelab/CorBinian . To investigate thermodynamic properties of neural population codes , Tkačik et al [7] introduced a temperature parameter T for eq 5: P T ( x | λ ) = 1 Z T exp ( 1 T ( h ⊤ x + x ⊤ J x + ∑ k = 0 n V k δ ( K ( x ) = k ) ) ) ( 7 ) Model fits are obtained at T = 1 , and the temperature parameter T is scaled to study the system ( i . e . characterised by PT ( x|h , J , V ) for T = 1 ) . Varying T , in effect , modulates probabilities by exponentiating them with 1/T , PT ( x ) ∝ ( PT=1 ( x ) ) 1/T , ( 8 ) and that the family of probability distributions obtained by varying T can be constructed for any distribution , not just maximum entropy models . For large temperatures PT approaches a uniform distribution ( PT ( x ) ≈ 2−n for each x ) , whereas for small temperatures it converges to a singleton , PT ( x* ) ≈ 1 with x* = argmaxx ( PT = 1 ( x ) ) . The specific heat , as given in eq 2 , can be obtained from the variance of the log-probabilities of the model . As the variance in practice cannot be computed for large n , we obtained estimates of c ( T ) using a pairwise Gibbs sampler . The specific heat does not depend on ZT , as changing ZT results in a constant , additive shift in log-probabilities which does not affect the variance . We tracked the variance of log-probabilities over an MCMC chain x ( 1 ) , … , x ( M ˜ ) of length M ˜ sampled at temperature T , using c ( T ) ≈1n ( E^[ logPT ( x ( m ) |λ ) 2 ]−E^[ logPT ( x ( m ) |λ ) ]2 ) ( 9 ) where E ^ denotes the average over spike words x ( m ) sampled from the the MCMC chain . For each population , we evaluated c ( T ) for 31 temperatures between T = 0 . 8 and T = 2 , and found the Gibbs sampler to provide reliable estimates over this temperature range—we in particular chose the minimal temperature T = 0 . 8 larger than previous previously in [7] to minimize possible effects from the sampler getting stuck ( see e . g . [46] ) . We used a burn-in of 2 . 0e4 sweeps , and ran the sampler for ( n100 ) 2×4h of CPU time , resulting in between 9 . 97e5 and 1 . 72e6 sweeps for n = 100 ( i . e . between 4 . 94e9 and 8 . 52e9 sampled individual spike words ) . For the theoretical analysis of the sampling process , we adopted a class of population models ( here referred to as ‘flat’ models ) in which all neurons are drawn from an infinite pool of neurons which all have identical mean firing rates , pairwise correlations and higher-order correlations [44 , 25 , 70 , 3 , 71] . Such a model is fully specified by the population spike-count distribution P ( K = k ) , and all spike words with the same spike count are equally probable . As a result , the probabilities of individual patterns x can be read off from the spike-count distribution by P ( x ) = ( kn ) −1P ( K=k ) ( 10 ) whenever ∑ i = 1 n x i = k . In a maximum entropy formalism , this model can be obtained by setting hi = 0 and Jij = 0 for all i , j ∈ {1 , … , n} and only optimising entries of V . Without loss of generality , we fixed fixed V0 = 0 [43] , resulting in n degrees of freedom for the model . In flat models , it is possible to explicitly construct a limit n → ∞ which will help us understand population analyses performed on experimental data: We assume that there is a spike-count density f ( r ) , r ∈ [0 , 1] , which describes the population spike-count distribution of an infinitely large population . f ( r ) denotes the probability density of a fraction of r neurons spiking simultaneously . Finite-size populations of n cells are then obtained as random subsamples out of this infinitely large system . Based on previous findings by [25] , we show in section 2 . 3 in S1 Supporting Information that , in this construction , flat models always exhibit a linear divergence of specific heat , unless the limit f ( r ) is given by either a single delta peak or a mixture of two symmetric delta peaks . These two models corresponds to systems that ( for large n ) either behave like a fully independent population ( whose spike-count distribution converges to a single delta peak ) , or a population described by a pure pairwise maximum entropy model ( which converges to two delta peaks ) . In particular , any flat model with higher-order correlations [17 , 70 , 71] , or a non-degenerate f ( r ) , will exhibit ‘signatures of criticality’ . Furthermore , we show that , for continuous f ( r ) , c ( T ) does not diverge for any T ≠ 1 . In combination , these results show that the peak of the specific heat is mathematically bound to converge to T = 1 for n → ∞ in this model class . We further simplified the flat model by re-parametrising P ( K = k ) by a beta-binomial distribution , thereby reducing the number of parameters from n to two , and—importantly—obtaining parameters which do not explicitly depend on n . In this model , P ( K=k ) = ( kn ) Beta ( α+k , β+n−k ) Beta ( α , β ) = ( kn ) ∫f ( r ) rk ( 1−r ) n−kdr ( 11 ) and f ( r ) = 1 Beta ( α , β ) r α - 1 ( 1 - r ) β - 1 . ( 12 ) For simulated data , we found values for α , β extracted from the beta-binomial fits to populations of different sizes n to be stable over a large range of n ( Fig 3b ) . We used the beta-binomial parameters obtained from the largest investigated n to estimate the divergence rate c ˜ for n → ∞ .
|
Understanding how populations of neurons collectively encode sensory information is one of the central goals of computational neuroscience . In physics , systems are often characterized by identifying and describing critical points ( e . g . the transition between two states of matter ) . The success of this approach has inspired a series of studies to search for analogous phenomena in nervous systems , and has lead to the hypothesis that these might be optimized to be poised at ‘thermodynamic critical points’ . However , translating concepts from thermodynamics to neural data analysis has been a challenging endeavour . We here study the data analysis approaches that have been used to provide evidence for criticality in the brain . We find that observing signatures of criticality is closely linked to observing activity correlations between neurons– a ubiquitous phenomenon in neural data . Our study questions the experimental evidence that neural systems are optimised to exhibit thermodynamic critical behaviour . Finally , we provide practical , open-source tools for analyzing large-scale measurements of neural population activity using maximum entropy models .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"applied",
"mathematics",
"ocular",
"anatomy",
"neuroscience",
"simulation",
"and",
"modeling",
"algorithms",
"probability",
"distribution",
"mathematics",
"ganglion",
"cells",
"computational",
"neuroscience",
"thermodynamics",
"coding",
"mechanisms",
"research",
"and",
"analysis",
"methods",
"animal",
"cells",
"probability",
"theory",
"physics",
"cellular",
"neuroscience",
"retina",
"cell",
"biology",
"anatomy",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"afferent",
"neurons",
"physical",
"sciences",
"ocular",
"system",
"retinal",
"ganglion",
"cells",
"computational",
"biology"
] |
2017
|
Signatures of criticality arise from random subsampling in simple population models
|
Using protein counts sampled from single cell proteomics distributions to constrain fluxes through a genome-scale model of metabolism , Population flux balance analysis ( Population FBA ) successfully described metabolic heterogeneity in a population of independent Escherichia coli cells growing in a defined medium . We extend the methodology to account for correlations in protein expression arising from the co-regulation of genes and apply it to study the growth of independent Saccharomyces cerevisiae cells in two different growth media . We find the partitioning of flux between fermentation and respiration predicted by our model agrees with recent 13C fluxomics experiments , and that our model largely recovers the Crabtree effect ( the experimentally known bias among certain yeast species toward fermentation with the production of ethanol even in the presence of oxygen ) , while FBA without proteomics constraints predicts respirative metabolism almost exclusively . The comparisons to the 13C study showed improvement upon inclusion of the correlations and motivated a technique to systematically identify inconsistent kinetic parameters in the literature . The minor secretion fluxes for glycerol and acetate are underestimated by our method , which indicate a need for further refinements to the metabolic model . For yeast cells grown in synthetic defined ( SD ) medium , the calculated broad distribution of growth rates matches experimental observations from single cell studies , and we characterize several metabolic phenotypes within our modeled populations that make use of diverse pathways . Fast growing yeast cells are predicted to perform significant amount of respiration , use serine-glycine cycle and produce ethanol in mitochondria as opposed to slow growing cells . We use a genetic algorithm to determine the proteomics constraints necessary to reproduce the growth rate distributions seen experimentally . We find that a core set of 51 constraints are essential but that additional constraints are still necessary to recover the observed growth rate distribution in SD medium .
A cell’s phenotype—its set of distinguishing observable traits—can be as much an emergent property of the cell’s environment and gene expression state as it is a result of the cell’s genotype . While some observables , like an organism’s response to Gram staining , can be immutable and tied to specific genes , others can be more fluid , varying from cell-to-cell with the random fluctuations in each cell’s molecular makeup [1–4] . A cell might by chance over- or under-express the enzymes involved in a given biosynthetic pathway , in which case the over- or underproduction of that pathway’s end product might signify a naturally occurring phenotype . Understanding this type of phenotypic variability requires models capable of connecting comprehensive gene expression profiles with cellular function . Constraint-based methods like flux balance analysis ( FBA ) have proven to be among the more successful approaches to modeling complex enzyme-mediated biochemistry at the cell scale ( for recent reviews and a primer , see [5–9] ) . In its simplest form , FBA seeks the flux distribution through a biochemical network that maximize the production of some specific objective , like biomass , while requiring that the concentrations of all other metabolites remain fixed ( i . e . the flux into and out of each metabolite is balanced ) . Parsimonious FBA ( pFBA ) improves on the predicted flux distribution [10] by minimizing the total flux through all reactions while maintaining optimal objective function . Minimizing total flux reduces the number of feasible flux distributions and represents efficient enzyme usage by the cell . By imposing constraints on the flux allowable through certain reactions ( such as substrate uptake reactions , or reactions catalyzed by mutated , knocked-out , or low-copy number enzymes ) , different environments , genetic perturbations , or gene expression states can be modeled . The use of FBA and related techniques has grown to include a large user-base that actively contributes to the development of both methods and models , and metabolic reconstructions now exist for a variety of model organisms ranging from bacteria and yeast up through humans [11–15] . A particularly vibrant area of research in the field has been the use of large -omics data sets to constrain models in ways that reflect the influence of the cell’s regulatory machinery . RNA microarray and RNA-Seq data can be used to impose reaction constraints according to the expression levels of the genes that encode their associated enzymes [16–20] . More recently the development of coupled metabolism and expression ( ME ) models has allowed for the direct prediction of the enzyme expression state that optimizes growth , yielding results that agree with experimental data sets [21] . While these methods yield insight into the average behavior of a population , they say little about cell-to-cell variability among sub-populations . Heterogeneity in gene expression has been the subject of intense experimental and theoretical research over the past several years [22–32] , but relatively few studies have attempted to understand its effects on cellular function [33 , 34] . Gene expression is known to correlate with growth rate [35]; Labhsetwar et al . [34] developed the Population FBA methodology ( Fig 1 ) in order to show that by sampling experimentally determined enzyme copy number distributions in a correlated fashion and using them as constraints on a genome-scale model of Escherichia coli metabolism , independently simulated cells exhibit a broad distribution of growth rates and several behavioral phenotypes ( e . g . some cells secrete acetate while others do not , or some cells make heavy use of the Entner-Doudoroff pathway while others predominantly use the Embden-Meyerhof-Parnas pathway ) . This study was made possible by the results of Taniguchi et . al . ’s groundbreaking fluorescence microscopy investigation of protein expression in single cells with single molecule sensitivity [36] . With recent developments in microscopy and microfluidics , a number of research teams have begun to report direct observations of single-cell growth rates [37–39] . Intriguingly , growth rate distributions reported in yeast [38] bear a striking resemblance to that predicted in the Labhsetwar et . al . , article; in particular both show a broad “shoulder” of slow-growing cells and a distinctive peak of fast-growers ( see Fig 2 ) [34 , 38] . Here we extend the Population FBA approach developed in [34] in order to predict and characterize emergent metabolic phenotypes within yeast populations . Employing the yeast 7 . 6 metabolic reconstruction—the latest and most complete and predictive genome-scale metabolic reconstruction of Saccharomyces cerevisiae to date [14 , 41]—along with comprehensive proteomics [42] and microarray [43] data sets , we construct highly realistic populations of independent in silico S . cerevisiae cells in two different growth media ( the first , denoted SD represents the same synthetic defined medium used in [42] , while the second , denoted 13C , represents the minimal medium used in a recent experimental fluxomics study [40] , See Table A1 in S1 Text ) . By ensuring these populations realize experimentally observed growth-rates , levels of gene expression , and correlations among co-regulated genes , we are able to create detailed models of the intracellular metabolic fluxes of every individual cell ( ∼ 100 , 000 ) . We show that the Population FBA methodology using scaled protein counts and the yeast 7 . 6 metabolic reconstruction give quantitative and qualitative agreement with experimentally observed intracellular fluxes ( as determined by a 13C study [40] ) . The use of transcriptomics data in order to impose correlations among co-regulated genes marginally improves the fidelity of our predicted intracellular fluxes . We then characterize the dominant metabolic phenotypes within our modeled populations . Specifically we find a shift in the balance between fermentation and respiration among fast-growing cells , we find cells in amino acid-rich media that make use of a complex set of reactions involving the glycine cleavage system , we find cells in minimal media that leverage the pentose phosphate pathway in order to conserve NADH , and we find slow-growing cells whose uptake of certain amino acids from the media exhibits a distinctive bimodality . And finally , we characterize the degeneracy of the possible sets of enzyme-related constraints that can give rise to the experimentally observed growth rate distributions .
The consensus yeast metabolic model version 7 . 6 [14] was chosen to describe the metabolic pathways in our simulations . This model ( available from yeast . sourceforge . net ) represents the most comprehensive yeast metabolic reconstruction to date [41] . All FBA calculations were performed using the COBRA toolbox version 2 . 0 [44] or COBRApy [45] . Gurobi 6 was used to perform all linear programming optimizations . Flux variability analysis ( FVA ) was performed using COBRA function fluxVariability ( ) to determine robustness ( minimum and maximum ) in flux values given a percentage optimality . FVA was also used to identify proteins with significant copy numbers but zero predicted flux in their associated reactions . For every cell in our modeled populations we sampled fluorescence values out of 535 experimentally determined distributions and converted them to enzyme copy number using Eq 1 ( see Methods Section Conversion of fluorescence to protein copy numbers and scaling ) . Each sampled enzyme copy number was paired with a turnover rate corresponding to that enzyme’s function ( kcat ) , and the product of these and a conversion factor yielded the upper bounds for the fluxes through the reactions catalyzed by each enzyme in each cell ( See Methods Section Constraint relaxation for realistic growth ) . The conversion factor used was 3 . 0 × 10−7 s cell−1 mmol gDwt−1 hr−1 , given by the number of seconds in an hour ( 3 , 600 ) divided by the average dry mass of a haploid yeast cell ( 2 . 0 × 10−11 g [46] ) and the number of particles in a mmol ( 6 . 02 × 1020 ) . In cases where multiple enzymes catalyze a given reaction , Gene-Protein-Reaction ( GPR , part of the metabolic model ) rules were used to determine the effective upper bound for the reaction from the upper bounds calculated for the individual enzymes . In cases involving “AND” relationships ( i . e . an enzyme is made up of two subunits and both need to be present ) , the minimum of the individual upper bounds was used , whereas in cases involving “OR” relationships ( i . e . different proteins can catalyze the same reaction ) , the sum of the individual upper bounds was used . If a count was missing for one of the enzymes involved in an “OR” relationship , the upper bound was left at the default value of 1 , 000 mmol gDwt−1 hr−1 . After setting all protein-associated constraints , parsimonious FBA [10] was performed in order to predict the internal fluxes of each modeled cell . Upper bounds for the uptake substrates were applied depending on the growth medium being modeled . The SD medium included glucose , 19 amino acids , uracil , citrate , vitamins , and minerals; the upper bounds for the amino acids , uracil , citrate and the vitamins were estimated based on experimental data [47] ( when no data was available the maximum experimental uptake was set , See Table A1 in S1 Text ) , those for oxygen and the minerals were unconstrained , and glucose upper bound was scaled to match experiment [40] . The strain , BY4741 , used in the growth rate distribution [38] and proteomics [42] studies—both grown in SD medium—contained several gene deletions , including his3Δ1 , leu2Δ0 , met15Δ0 , and ura3Δ0 . To account for this , the genes YCL018W , YLR303W , YEL021W were inactivated , leading to zero flux being allowed through five reactions: 3-isopropylmalate dehydrogenase ( r_0061 ) , cysteine synthase ( r_0312 ) , O-acetylhomoserine ( thiol ) -lyase ( r_0812 , and r_0813 ) and orotidine-5-phosphate decarboxylase ( r_0821 ) . The histidine biosynthesis knockout is recovered when GFP is tagged to any protein , so the gene YOR202W was kept active . The 13C medium included only glucose , some vitamins , and minerals ( See Table A1 in S1 Text ) . As in the SD medium , vitamin uptake upper bounds were set based on experimental data [47] , the oxygen and minerals were unconstrained and glucose upper bound was scaled to match experiment [40] . Glucose uptake upper bound of 20 mmol gDwt−1 hr−1 is also supported by Diderich et al . [48] . The strain used in the 13C medium , FY4 Mat a , is a wild-type strain , so no modifications were done to original yeast 7 . 6 model to simulate this . Protein abundances were obtained from single cell fluorescence measurements on yeast strain BY4741 grown on glucose SD medium [42] . The authors reported fluorescence distributions that were calculated from average pixel intensities over entire cells; we therefore considered all protein abundances to be size-normalized . For each GFP-labeled protein , Dénervaud et al . [42] deconvoluted the single cell fluorescence signal from the autofluorescence signal , and fitted the results to gamma distributions , providing shape and scale parameters for 4 , 159 proteins measured at 40 time points , taken 20 minutes apart ( totaling 166 , 360 fluorescence distributions ) . Since the study aimed at observing changes in the proteome in response to stress , only 18 of the 40 time steps could be used for each protein ( the ones before the induction of stress factors ) . Of the 74 , 862 remaining distributions , several displayed significant abnormalities , most likely resultant from the automated deconvolution procedure used to separate weak GFP fluorescence signals from the cell’s autofluorescence . The abnormality consisted of fluorescence distributions that were extremely narrow and usually had low mean fluorescence , hereafter referred to as “spikes” . Examples can be seen in Fig 3 . A multi-step procedure was developed to automate the processing of the almost 75 thousand fluorescence distributions , and when appropriate , censoring of spikey distributions . Only proteins that had data for all 18 time points were subjected to this process , which led to the removal of 59 proteins . First , a conservative lower bound of 0 . 1 was placed on standard deviations to remove the most obvious spikes , leading to the removal of all 18 time points for 7 proteins and a total of 2948 total fluorescence distributions being discarded across all proteins . Then , for each protein , the remaining distributions were used to determine a central reliable region for means and standard deviations , which were defined as the range from 1 . 5 times their IQR ( inter quartile range ) below the 25% quantile to 1 . 5 times the IQR above the 75% quantile . If a fluorescence distribution had either its mean or standard deviation outside this range , the distribution was discarded , leading to the removal of another 4175 distributions . After this step , only proteins that had 3 or more fluorescence distributions out of the original 18 were kept , which lead to the removal of another 3 proteins . Finally , the third step calculated the coefficient of variation ( CV , defied as standard deviation over the mean ) of the means and the CV of the standard deviation of the remaining distributions for all proteins . Only proteins whose distributions had both means and standard deviations with CVs lower than an upper bound of 0 . 5 were kept , removing 201 additional proteins . Proteins with mean fluorescence less than 7 . 98 A . U . were also removed because Dénervaud et al . [42] considered them unreliable . We found that these proteins had significantly less noise than proteins with means higher than 7 . 98 A . U . The final set of reliable fluorescence distributions represented a total of 3 , 647 proteins , which covered diverse cellular processes and compartments . The full dataset acquired after this process is reported in the S2 File , including parameters for fluorescence distributions in individual timesteps , and full plots for all fluorescence distributions used in our simulations . The fluorescence distributions which were found to be reliable were then converted to absolute protein copy distributions . We used single cell quantification of 10 proteins ( Table 1 ) from mass spectrometry ( MS ) [49] in order to relate fluorescence values to single cell copy numbers ( Fig 3 ) . The quantitative protein abundance from the MS study were determined using the same yeast strain as used in Dénervaud et al . , but were grown on complex media . In order to estimate protein counts for synthetic defined ( SD ) media , we used expression ratios observed in a single cell proteomics study [50] , where protein abundances were measured in both complex and synthetic defined media . Finally , a linear fit between log values for protein counts and fluorescence was used to obtain the Eq ( 1 ) for converting fluorescence into protein counts: p = 2 . 87 * f 1 . 5577 ( 1 ) where p represents the single cell protein copy number , and f represents the fluorescence value . During sampling , we ensured a lower bound of 2 . 87 for all enzymes ( if a copy number was sampled lower , it was replaced with 2 . 87 ) , This was because we expect fluorescence values less than 1 to be unreliable . Protein counts were scaled in case of simulation for 13C medium because the protein distributions measured by Dénervaud et al . are in SD medium . Ratios to scale the protein counts were found from microarrays comparing gene expression of cells grown in SD medium to cells grown in SD medium without amino acids for 6 hours after being transferred from SD medium [51] . Top 10 proteins downregulated and top 10 proteins upregulated in minimal medium are shown in Table A2 in S1 Text . Microarray datasets from Kemmeren et . al . [43] ( available from the GEO database , Accession No . GSE42528 ) , were used to calculate correlation coefficients among the 532 out of 535 metabolic proteins we sampled . Rest of the three proteins were not measured in these microarray experiments . This microarray data is well-suited for our study because it was produced using an almost identical strain of yeast , BY4742 which has same deletions but different mating type , and under similar growth conditions as that used in the proteomics study we rely on [42] . Absolute fluorescence values for the sample channel of the two-channel microarrays were used; because almost all of the genes evaluated had two probes on the microarray chip ( Accession no . GPL11232 ) , the mean value of the two probes was computed . Fluorescence values were then quantile normalized across the entire set of microarray data [52] . Correlation coefficients were calculated from these normalized fluorescence values . These correlation coefficients were then used to create correlated samples of protein counts using the usual Cholesky decomposition methodology [34] . The correlations observed show clear biological relevance . For example , the Crabtree effect , which is well known in S . cerevisiae , can be seen in the positive correlations among the Glucose transporter HXT1 and genes in the fermentative pathway as well as the negative correlations between HXT1 and genes involved in the TCA cycle and oxidative phosphorylation ( see Fig A11 in S1 Text ) . Moreover , the correlations we see recover many experimentally known regulatory links in yeast ( see Fig A12 and Section Reliability of mRNA Microarray Correlation Data in S1 Text ) . Without internal constraints , the metabolic model iJO1366 for E . coli and the yeast version 7 . 6 model both return higher growth rates for a given glucose uptake rate than is experimentally observed . However , as previously reported in the population studies on E . coli [34] , imposing all of the possible constraints arising from the measured protein distributions and turnover rates does not allow the population to grow . The problem lies with either some of the protein counts or some of the turnover rates . In converting the fluorescence data to protein distributions , we already removed spurious data and low counts , so we were confident in the remaining distributions . Moreover , because a third of the kcat values obtained from BRENDA have changed in the span of a year , we chose to keep the protein counts and raise the appropriate turnover rates in order to allow for growth . To deterministically find problematic turnover rates , we iteratively simulated populations of 400 cells and identified the reaction whose flux most often reached its imposed upper bound . kcat for the enzyme associated with that reaction was doubled . If that reaction was catalyzed by multiple proteins , we doubled the kcat value for the protein with highest protein mean count in case of isozymes ( ‘OR’ relationship ) and all the subunits in case of an protein enzyme complex ( ‘AND’ relationship ) . We continued this procedure iteratively until the mean growth rate of the sampled population reached 0 . 35 hr−1 , the bulk growth rate measured in both the proteomics [42] and single cell ( microcolony ) experiments [38] . First round of doublings helped us to focus on proteins which needed excessive doublings and hence manual search for those kcat values was performed and any higher kcat reported in literature was accepted . Manually found kcat values can be found in Table A3 in S1 Text . Before going through the doubling procedure , we also raised kcat values of all subunits in a protein complex to the highest kcat among the subunits . Each yeast cell in our modeled populations had a unique protein copy number for 535 genes , and a unique flux distribution throughout the metabolic network of over 3 , 400 reactions . Different fluxes in this metabolic network are linearly dependent on each other and constitute metabolic pathways . To find pathways that were differentially used by different segments of our modeled populations , we used principal component analysis ( PCA ) as implemented in MATLAB’s pca ( ) function to elucidate orthogonal directions ( in the 3 , 400-dimensional flux-space ) in which the cells in our populations varied most . We chose 1 , 000 cells at random from the population for this analysis . Since the members of this population grew at different growth rates , we normalized all fluxes by the cell’s growth rate , allowing us to identify growth-independent differences in pathway usage . This methodology is similar to that used previously [34] , but we didn’t need to rotate the components coming out of PCA as they aligned with canonical metabolic pathways . A new procedure for filtering overly-constraining turnover rates based on the Micro Genetic Algorithm ( GA ) formalism was developed [53] . This method utilizes an entire growth distribution as a target for optimizing the selection of experimental constraints . Micro Genetic Algorithm was chosen instead of a “regular” Genetic Algorithm solely for computational cost concerns . In a “regular” GA algorithm in dozens to hundreds of genomes would have to be simulated at each generation , and several hundred generations could need to be evaluated to reach the same results . The computational cost would be extremely higher as compare to our GA implementation . In our attempt to reduce the size of search space we have restricted GA variables to binary values representing weather to use a particular kcat or 38 , 000 s−1 rather than more flexible values kcat can take in the doubling procedure . Briefly , a population of 10 “genomes” was simulated , each one composed of a list of “genes” that indicated if a protein’s kcat would be kept at its BRENDA value , or if it would be raised to 38 , 000 s−1 . The genomes were allowed to evolve by exchanging information , and each new generation was created by a random selection of solutions biased by their fitness , while always taking the best solution to the next generation ( see SI Section Extended Methodology: Genetic Algorithm for Constraint Selection for details ) . The fitness of each genome was determined by simulating a cell population based in its kcat selection , and then calculating the goodness-of-fit between the resulting growth rate distribution and the observed distribution [38] .
The basic Population FBA methodology has been described previously [34] . Briefly , enzyme copy numbers are sampled from experimentally-determined distributions [42] from a single cell proteomics study; each sampled set of enzyme copy numbers represents a unique cell in its own gene expression state . Assuming Michaelis-Menten kinetics , each copy number—paired with an appropriate enzyme turnover rate ( kcat ) —represents the maximum reaction flux that the cell can maintain through the reaction ( s ) mediated by that enzyme . Many genes are known to exhibit some correlation in their expression levels . For bacteria , [34] , this effect was handled fairly simply; proteins in the “extrinsic noise limit” , noise floor observed in proteins with high means , were assumed to exhibit a correlation coefficient of 0 . 66 suggested by the single cell proteomics study [36] . Due to the availability of large transcriptomics datasets , we are now able to take a more refined approach in which we systematically impose the types of correlations that should naturally arise among the copy numbers of co-regulated proteins . This was accomplished by extracting correlation coefficients for ∼ 4 , 000 S . cerevisiae gene products from an expansive collection of microarray gene expression datasets [43] and using them to draw correlated sets of protein copy numbers Constraints of this type were then imposed throughout a genome scale flux balance model of metabolism , and parsimonious flux balance analysis ( pFBA ) [10] was used to predict each cell’s metabolic behavior . The copy number distributions that were used were adapted from a recent article by Dénervaud et al . [42] for yeast grown in SD medium . The authors used a GFP fusion library spanning 4 , 159 S . cerevisiae proteins and a unique parallel microchemostat microfluidic device to measure single cell fluorescence intensity distributions—representative of protein expression distributions—for approximately 2/3 of the yeast proteome . Intensities sampled from these distributions were transformed to copy numbers using a calibration curve ( see Fig 3A ) . Several of the measured fluorescence distributions were abnormally “spikey , ” likely as a result of poor deconvolution of the GFP signal and the cell’s own autofluorescence ( see Fig 3B and 3C ) . We removed these spikey distributions by determining which distributions had abnormal means and standard deviations using a simple outlier-detection protocol ( see Methods Section Conversion of fluorescence to protein copy numbers and scaling ) . Among the remaining 3 , 885 distributions , those with mean fluorescence lower than 7 . 98 A . U . ( as measured by Dénervaud et al . [42] ) had significantly lower noise than the proteins with similar means hence they were also removed . The 3 , 647 distributions that remained after this censoring procedure showed noise characteristics that agreed qualitatively with previously published results in E . coli ( see Fig A1 in S1 Text ) . Only 535 of these remaining distributions were associated with enzymes involved in the yeast 7 . 6 metabolic reconstruction [14] ( see S1 File ) , and thus only these were used in our study . We would like to note that GFP is extremely stable protein which might affect stability of tagged protein and hence bias the protein counts towards higher number than their numbers in untagged cells . Metabolic reconstruction of yeast accounts for 13 compartments which represent various organelles and their membranes . All metabolites are assigned to one of these compartments and reactions are either localized in a compartment if all the reactants and products are present in the compartment or facilitate transport of metabolites across compartments . When we associate an enzyme with a reaction using the gene-protein-reaction associations of the reconstructions , we assume all the copies of the enzyme are available to the reaction it is associated with . So even though the copies of the enzyme might be spread out over multiple compartments in real cells , in lack of that information we make all the copies available to all the reaction the enzyme is associated with . Two sets of simulations were performed , corresponding to the two different environmental conditions . The first was intended to replicate the cell growth media used in a study ( [40] ) of 13C-labeled glucose utilization by several strains of yeast . This was done in order to accurately compare our predicted intracellular metabolic fluxes with those determined experimentally ( see Results Section Population FBA yields intracellular fluxes that agree with 13C fluxomics data ) . The synthetic defined ( SD ) medium replicates the conditions used in single cell proteomics [42] and growth rate distribution studies [38] . This SD media included approximately the same concentrations of salts , double the glucose , and several metabolites not present in the 13C media . These included 19 amino acids ( including the histidine , leucine , and methionine necessary for the growth of the his3Δ1 , leu2Δ0 , and met15Δ0 experimental strain ) , as well as citrate , and uracil ( necessary for the ura3Δ0 also present in the experimental strain ) . Our modeled cells contained the same knockouts as the cells used in the experiments . Modeling of 13C media involved modifying relevant uptake rates in the metabolic model and scaling protein copy numbers measured in SD media to 13C media . Details of both modeled media and rescaling can be found in Section Extended Methods: Metabolic Model and Experimental Data and Table A1 and Table A2 in S1 Text . We assume that the relative composition of biomass is the same in all members of the population , although some experiments indicate the composition may change as a function of the growth rate [54 , 55] . FBA models in general are underdetermined . By adding constraints in the form of reaction upper- and lower-bounds , modelers are able to whittle down the solution space ( the right null space of the stoichiometry matrix ) to the flux distributions that most accurately describe real cells ( [7] ) . Metabolic reconstructions already include topological and thermodynamic constraints in terms of stoichiometric matrix and reaction reversibilities . Additional constraints are also routinely added to reflect the genetics of the strain ( for example by fixing the flux through a reaction mediated by a “knocked-out” gene to zero ) as well as the growth medium used ( for example , by limiting the uptake of substrates absent from the media to zero ) . In Population FBA we add additional constraints based on protein copy numbers and their kinetic capacity . After censoring proteomics data to remove unreliable distributions , we believe we have good quality of protein copy number distributions . As for the kinetic capacity , we rely primarily on the BRENDA database [56 , 57] . The BRENDA database often contains several sets of kinetic parameters for a given reaction . These can include values for enzymes from different organisms , strains , and most often in vitro conditions . A recent study concluded that kcat measured in vitro generally agree with max kcat in vivo estimated using omics data [58] . Whenever possible the largest kcat value available for a wild-type S . cerevisiae strain was taken , otherwise the largest value reported for any mutant or other species was used . If no kcat was available for an enzyme-mediated reaction , a value of 38 , 000s−1 ( corresponding to the largest kcat reported for a wild-type yeast enzyme in BRENDA ) was set . These criteria were adopted in order to minimally constrain the model . Importantly , the 535 sampled enzymes and kcat values could in principle be used to impose constraints on 1 , 128 of the model’s 3 , 493 reactions ( each enzyme catalyzes two reactions on average ) , but it was found that imposing all of these constraints impedes the growth of the modeled cells to levels well below that seen experimentally . Several enzyme turnover rates were found to have published values well below those necessary to allow realistic growth . For example , phosphofructokinase ( PFK ) , which is made up of two subunits , had mean copy numbers measured to be 103 , 880 ( α subunit ) and 47 , 919 ( β subunit ) and a reported turnover rate of 62 s−1 ( See Table 2 ) . This led to a maximum reaction flux of 1 . 16 mmol gDwt−1 hr−1—approximately ten-fold smaller than the experimentally measured 13C glycolytic flux [40] . In cases like this , the kcat values , which are relatively uncertain ( reported values for phosphofructokinase , for example , range over four orders of magnitude [57] ) , were “doubled” until the mean population growth rate of 0 . 35 hr−1 was achieved ( see Methods Section Constraint relaxation for realistic growth for details ) . Our doubling methodology involves iteratively generating small populations of modeled cells ( 400 ) and then determining which reaction most constrains cellular growth , and doubling its kcat . This strategy revealed that certain enzymes required excessive numbers of doublings ( for example , the kcat for Glycogen Synthase was doubled 19 times in our 13C simulations ) . These rates were investigated further , and in many cases we were able to find significantly higher kcat values in the literature than were reported in BRENDA ( See Table A3 in S1 Text ) . Even after including kcat values from literature , some protein’s kcat needed significant doubling e . g Acetylornithine aminotransferase needed 13 doublings . We also found kcat for NAD dependent methylenetetrahydrafolate dehydrogenase ( YKR080W ) , was wrongly listed in BRENDA as 1 . 63 s−1 and needed 12 doublings in simulation for SD medium . The correct kcat is 1 , 643 s−1 [59] which is much closer to the value required to sustain flux in SD medium ( 6 , 676 s−1 , Table A3 in S1 Text ) and would be obtained after only 2 doublings . We note that approximately 1/3 of the kcat values taken from the BRENDA database have changed within just the past year , further casting doubt on the reliability of these parameters and suggesting the need for further consistency checks among the metabolic fluxes . In both modeled environments , overly-constraining enzyme turnover rates were doubled until the mean population growth rate matched that reported in the respective experimental studies . In total , 391 doublings were performed which affected 121 kcat values in order to match the SD media growth rate of 0 . 34 hr−1 reported in [42] , while 506 doublings were performed which affected 146 kcat values to match the 13C media growth rate of 0 . 35 hr−1 , reported in [40] ) . In each case , the resulting population exhibited a spectrum of specific growth rates that ranged from nearly 0 . 0 to over ∼ 0 . 57 hr−1 . Importantly , although only the mean growth rate was fit , the distribution for the SD medium was nearly identical to the experimentally measured distribution [38] shown in Fig 2b . While a mean growth rate was reported , the corresponding growth distribution curve is not available for cells grown in 13C medium . Sampling from protein distributions that have been rescaled for 13C medium allows us to predict such a curve ( See Fig A3 in S1 Text ) . Both modeled populations exhibited the same broad shoulder of slow-growers , and prominent peak of fast-growing cells seen in experiments . The fast-growing peak was predominantly the result of limitation in glucose uptake ( although limitation in certain amino acids also contributed in the SD media ) . While slow-growing cells utilized glucose at rates below the maximum allowed rate; see Fig A5 and Fig A8 in S1 Text ) , the fast-growers did tend to reach their limit . This common glucose limitation resulted in most of these cells having very similar growth rates , which in turn resulted in the pronounced peak we see . The broader tail of slow-growing cells was the result of important enzymes being sampled at low copy numbers . For example , sampled upper bounds on the ATP synthase and ubiquinol-ferricytochrome c reductase reactions , which participate in respiration , limited the growth of approximately 50% of the slow-growing subpopulation ( those growing at rates less than 0 . 34 hr−1 , see Fig A9 and Fig A6 in S1 Text ) . A similar glucose-associated peak arose in work modeling the growth rate distribution of E . coli [34] . In light of this and the agreement we see between our simulated distributions and the experimental one , we wonder if substrate ( e . g . glucose , amino acids ) limitation could in general lead to similar peaked growth rate distribution across a range of organisms . It should be straightforward to investigate this experimentally . In the simplest case , growth rate distribution experiments similar to that of [42] could be carried out using media with varying concentrations of substrates . In a similar vein , the contribution of individual substrates to the peak could be investigated in strains in which the transporters of a substrate are expressed under the control of an inducible regulatory element . In E . coli , for example , the glucose transporter ptsG might be expressed under the control of the lac system—the shape and location of any peak in the growth rate distributions of cells grown with and without IPTG ( a lactose analogue used in this case to induce ptsG expression ) could then indicate whether glucose uptake limitation is responsible . In order to identify the features of protein distribution which play important role in obtaining the shape of the distribution , we performed some control simulations . We randomly omitted 50% ( 267 ) and 66% ( 356 ) of the 535 available protein distributions and used the rest of the protein distributions to perform Population FBA . The resulting growth rate distributions are very similar to experimental growth rate distributions with a broad tail at slow growth rates and a peak at fast growth rates ( See Fig A14 in S1 Text ) . In our population of 100 , 000 cells , only 165 out of the 3 , 493 reactions are ever constrained which is indicative of few proteins giving the shape to the growth rate distribution . We also tried to change all the protein distributions from gamma distribution to either normal or uniform distributions while keeping the mean intact . Resulting growth rate distribution had no resemblance to the experimental growth rate distribution ( Fig A15 in S1 Text ) emphasizing the importance of the shape of protein distributions . We have seen this before with E . coli [34] that very few proteins constraint the growth of majority of cells in the population . Flux balance models can be evaluated across several different measures of predictiveness . Among the more common is the model’s ability to discern lethal and non-lethal gene knockouts . While this type of information can have important implications in synthetic biology and bioengineering [60–62] , it remains a somewhat blunt measure of a model’s overall utility . The yeast 7 model was recently shown to accurately predict gene essentiality [41] , but this does not necessarily translate into accurate network fluxes or realistic growth rates . Because our current work deals with the ability of our Population FBA methodology to predict pathway usage and growth rates , the most convincing test of our predictions is to compare them directly to experimental network flux measurements . Detailed measurements of intracellular metabolic fluxes are relatively scarce in the literature , owing in part to the cost and difficulty of such experiments . Nevertheless , a recent study used 13C-labeled glucose and some modeling to characterize eight fluxes involved in the central-metabolism of seven species of yeast [40] . These fluxes included glucose uptake , phosphogluconate dehydrogenase ( in the pentose phosphate pathway ) , fructose bisphosphate aldolase ( FBA ) , citrate synthase ( CS ) , malate dehydrogenase ( MDH ) and the production of glycerol , ethanol , and acetate given in Table 2 and Fig 4 ( for a full list of reaction names and abbreviations , see Table A4 in S1 Text ) . The authors found that S . cerevisiae ( a “Crabtree-positive” yeast ) respired little even under highly aerobic conditions . Using the wild-type yeast 7 . 6 model in 13C medium calculations were performed using our Population FBA methodology ( see Methods Sections Model , software , and Population FBA methodology and Constraint relaxation for realistic growth , and Fig A3 , A4 and A5 in S1 Text ) . The central metabolic fluxes of 1 , 000 simulated cells were compared to the experimental values . Comparison of these results with those predicted using FBA without proteomics constraints ( Fig A2 in S1 Text ) and our Population FBA methodology showed two important findings: 1 ) our Population FBA method imposes internal flux constraints in a manner that recovers the experimentally observed Crabtree effect , and yields mean network fluxes that by and large agree with experiment ( with the exception of some underestimation of flux into the pentose phosphate pathway , and glycerol and acetate formation pathways , Fig 4 ) ; and 2 ) FBA without proteomics constraints fails to predict the Crabtree effect in S . cerevisiae . Our modeled cells exhibited a range of different metabolic behaviors . In order to characterize the ways in which they tended to differentiate , we employed principal component analysis ( PCA ) in a manner similar to that described in [34] ( see also Methods Section Analysis of sub-populations ) . The first two PCA components combined accounted for almost 85–95% of the total variability in pathway usage ( Table A5 in S1 Text ) . Analysis of the loadings of the first component showed that it was associated with similar metabolic behaviors , namely a shift toward respirative metabolism by fast growing cells , regardless of environment . In SD growth medium the second PCA component was associated with the cell’s NADH/NADPH economy where fast growing cells employed Serine-Glycine cycle to generate NADH and NADPH . Finally , although not elucidated by PCA , a particularly compelling form of metabolic variability characterized by a bimodality in the utilization of certain amino acids was found to emerge among the slow-growing cells modeled in SD medium . Threonine , aspartate , asparagine , glutamate , and glutamine all showed little correlation with growth rate , but each was found to be either not taken up at all , or taken up at its maximal allowable rate ( see Fig A8 in S1 Text ) . Variability in population behavior predicted by Population FBA in 13C growth medium is mostly one-dimensional ( Table A5 in S1 Text ) , similar to the dimension identified in first component of SD medium population , where movement towards higher growth rate implies restricted fermentation and increased respiration ( See Fig A4 in S1 Text ) . The main doubling algorithm ( see Methods Section Constraint relaxation for realistic growth ) we chose for raising overly-constraining kcat values focused on identifying constraints that most strongly limited cellular growth rate . This has the benefit of keeping as many of the experimental parameters available intact , but it also raises some questions , the most important of which is whether the set of turnover rates that are doubled is the only set that would lead to a mean growth rate similar to the measured one . Similarly , are all of the turnover rates that are kept intact after the doubling procedure actually necessary ? In order to address these questions , we implemented an Evolutionary Algorithm based approach for finding constraint sets that yield growth rate distributions that approximately recover the one seen experimentally . This method made use of the Micro Genetic Algorithm formalism , a type of genetic algorithm ( GA ) that uses relatively small numbers of “genomes” and dispenses entirely with “mutations” [53] ( see Methods Section Genetic algorithm for constraint selection ) . We performed 10 independent GA runs using the modeled SD media , and each resulted in a different set of turnover rates being lifted . In each case the resulting growth rate distribution closely matched the experimental one ( see Fig 9 ) . In general roughly twice as many turnover rates were lifted in each GA set as were affected by our main doubling method . This was because the GA associated no cost with filtering a value that did not impact the growth of the modeled cells . If , for example , a kcat constrained a reaction involved in metabolizing a sugar like galactose that was not available in the media , the main doubling method would never lift it because it would not constrain the growth rate , but the GA , being fundamentally a random search method , might lift it simply by chance . Across all GA runs , only 51 turnover rates were consistently raised ( see Table A6 in S1 Text and S1 File ) . Out of these 51 , 49 were also affected by doublings during our main doubling method , meaning that they represent a core set of problematic kcat values whose removal was necessary to achieve realistic growth . This core set was not by itself sufficient , however . In every GA run , between 150 and 181 additional turnover rates were also lifted . These extra kcat values showed little overlap among the 10 sets , which indicates that beyond the core 51 , the choice of which turnover rates to lift became highly degenerate . Despite this degeneracy , every set of turnover rate parameters found by the GA showed the same Crabtree effect and shift between fermentation and respiration yielded by our main doubling methodology ( see Results Sections Population FBA yields intracellular fluxes that agree with 13C fluxomics data , Fermentation vs . respiration and Fig 10 ) . As shown in Fig 11 , all 10 populations show usage of the serine-glycine cycle by fast growing cells similar to that shown in Figs 5 and 6 in the population obtained by doubling methodology SD media: Substrate level NAD+ Reduction , ADP phosphorylation , and organic acid efflux . Interestingly , the GAPDH-associated turnover rate that drove the bimodal amino acid utilization noted previously ( see Results Section SD media: Bimodality in amino acid utilization ) was lifted in some but not all of our GA runs . By comparing threonine usage across GA runs we found that when the GAPDH-associated kcat was raised the bimodality among slow-growing cells essentially disappeared; instead , at growth rates lower than about 0 . 3 hr−1 , the modeled cells all took up threonine at its basal rate and none were found to utilize the glycine cleavage system ( see Fig A10 in S1 Text ) . This finding further supports the notion that it is the GAPDH constraint that gives rise to bimodal amino acid utilization we observed . Among faster-growing cells , the glycine cleavage system , and the related uptake and catabolization of amino acids occurred regardless of which constraints were raised by the GAs ( Fig A10 in S1 Text ) . This is because glycolysis in the fast-growing cells is not constrained by enzyme copy-numbers , it is constrained by the glucose uptake rate itself; almost every cell growing faster that approximately 0 . 3 hr−1 experiences this limitation , and they engage in amino acid catabolism as a response . With the development of single cell and micro-colony imaging experiments [36–38 , 42] , instead of measuring a single growth rate ( via optical density , for example ) for an entire population , we can now observe a distribution of the growth rates of individual cells . To understand or interpret the general form of the growth rate distribution , we have to dig into the metabolic behavior of the underlying subpopulations . Recent systematic genome-wide fluorescence labeling studies have provided libraries of approximately 1 , 000 “strains” of labeled E . coli and 4 , 000 “strains” of labeled yeast . Examination of these strains has shown that proteins are not expressed at a specific number across a population . Due to the well-established innate stochasticity in essentially every cellular processes ( transcription , translation , DNA replication , cell division , etc . ) , these studies have shown that proteins are expressed in varying numbers from cell to cell . In order to understand how any given cell’s protein expression state effects its behavior , and how that behavior relates to the overall behavior at the population level , these protein distributions must be sampled and realistic subpopulations of individual cells must be modeled . Our Population FBA approach provides such a method; allowing us to carry out the generation of realistic populations of cells and subsequent analysis of their intracellular fluxes and exchanges with the environment . Simulations of the steady-state growth rates attainable by the cells in our modeled populations gives rise to a distribution that is in excellent agreement with the experimentally observed growth rate distribution [38] . In particular both show the same broad shoulder of slow growing cells ( ranging in growth rates from nearly 0 . 0 to approximately 0 . 3 hr−1 ) , and a dominant peak of fast-growing cells ( ranging between approximately 0 . 3 and 0 . 7 hr−1 , see Fig 2 ) . We show that substrate availability is the main cause of this peak , and we believe this can be verified experimentally; in particular we suggest micro-colony experiments similar to that of [42] under growth conditions with varying levels of substrate availability or with substrate transporters placed under the control of inducible regulatory elements ( see Results Section S . cerevisiae Exhibit a Broad Distribution of Growth Rates ) . Moreover , we find excellent agreement between experimental fluxomics data [40] and the computed intracellular fluxes predicted by our methodology , both within and between the cytosol and the mitochondria . These results underscore both the rigor of the Population FBA methodology as well as the high quality of the Yeast 7 . 6 metabolic model [14 , 41] . The simulations presented here also allowed us to make quantitative predictions about the effects of growth in media where the main difference was the presence or absence of amino acids . The 13C experiments on wild-type yeast contained no amino acids in the media; our simulations showed that cells under these conditions depended on ammonium , sulfate , and phosphate salts taken up from the media . The gene knockouts that differentiated the the strains used in the 13C and SD experiments required the addition of uracil , leucine and histidine to the SD media . The SD media also contained an additional 17 amino acids , several of which were taken up and catabolized as a energy source . Despite the differences , both simulated populations displayed very similar growth rate distributions . We have employed our Population FBA methodology to study metabolic heterogeneity in S . cerevisiae . One of the most important result of this study is that it underscores the need for imposing biologically realistic internal constraints in flux balance models . Without the types of constraints Population FBA imposes , the yeast 7 . 6 model gave fluxes , growth rates , and metabolic byproducts that were qualitatively and quantitatively inconsistent with the results of a 13C fluxomics study . Our study has shown that yeast populations exhibit the same types of cell-to-cell diversity in behavior that is coming to be recognized across the microbial world , and that although the particular sets of constraints that are necessary to recover the experimental growth rate distribution are not unique , any set that does recover the growth rate distribution also recovers the main metabolic behaviors we observed , including the Crabtree effect and the noted shift toward respiration seen among our fast-growing cells .
|
No two living cells are exactly the same . Even cells from a clonal population with identical genomes living in the same environment will express proteins in different numbers simply due to the random nature of the chemistry involved in gene expression . The consequences of this stochastic gene expression are complex and not well understood , especially at the level of large reaction networks like metabolism . Here we investigate how variability in the copy numbers of metabolic enzymes affects how individual cells extract nourishment from their environment and grow . We model 100 , 000 independent yeast cells , each with their own set of enzyme copy numbers sampled from experimental distributions , and use flux balance analysis ( FBA ) to compute the optimal way that each cell can use its metabolic pathways—an approach we dubbed Population FBA . We find that enzyme variability gives rise to a wide distribution of growth rates , and several metabolic phenotypes—subpopulations relying on diverse metabolic pathways . Most importantly , we compare the predicted fluxes through the different pathways to experimental values; we find that Population FBA is able to correctly predict Crabtree effect , while traditional FBA , which lacks the proteomics constraints our method imposes , differs both qualitatively and quantitatively from experiment .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"and",
"discussion"
] |
[
"cell",
"physiology",
"protein",
"metabolism",
"chemical",
"compounds",
"aliphatic",
"amino",
"acids",
"enzymes",
"enzymology",
"carbohydrates",
"cell",
"metabolism",
"organic",
"compounds",
"glucose",
"threonine",
"fungi",
"model",
"organisms",
"experimental",
"organism",
"systems",
"amino",
"acids",
"saccharomyces",
"research",
"and",
"analysis",
"methods",
"glycine",
"proteins",
"chemistry",
"yeast",
"biochemistry",
"eukaryota",
"cell",
"biology",
"organic",
"chemistry",
"monosaccharides",
"hydroxyl",
"amino",
"acids",
"biology",
"and",
"life",
"sciences",
"yeast",
"and",
"fungal",
"models",
"saccharomyces",
"cerevisiae",
"physical",
"sciences",
"metabolism",
"organisms"
] |
2017
|
Population FBA predicts metabolic phenotypes in yeast
|
The mean conditional fixation time of a mutant is an important measure of stochastic population dynamics , widely studied in ecology and evolution . Here , we investigate the effect of spatial randomness on the mean conditional fixation time of mutants in a constant population of cells , N . Specifically , we assume that fitness values of wild type cells and mutants at different locations come from given probability distributions and do not change in time . We study spatial arrangements of cells on regular graphs with different degrees , from the circle to the complete graph , and vary assumptions on the fitness probability distributions . Some examples include: identical probability distributions for wild types and mutants; cases when only one of the cell types has random fitness values while the other has deterministic fitness; and cases where the mutants are advantaged or disadvantaged . Using analytical calculations and stochastic numerical simulations , we find that randomness has a strong impact on fixation time . In the case of complete graphs , randomness accelerates mutant fixation for all population sizes , and in the case of circular graphs , randomness delays mutant fixation for N larger than a threshold value ( for small values of N , different behaviors are observed depending on the fitness distribution functions ) . These results emphasize fundamental differences in population dynamics under different assumptions on cell connectedness . They are explained by the existence of randomly occurring “dead zones” that can significantly delay fixation on networks with low connectivity; and by the existence of randomly occurring “lucky zones” that can facilitate fixation on networks of high connectivity . Results for death-birth and birth-death formulations of the Moran process , as well as for the ( haploid ) Wright Fisher model are presented .
Fixation is the replacement of an initially heterogeneous population with the offspring of just one individual . The probability of fixation and the average time that is required for a mutant to take over the population are two fundamental quantities in ecology and evolution . Both fixation probability and average fixation time have been widely studied by physicists and mathematicians for almost a century , starting with the early works by Haldane [1] , Fisher [2] , Wright [3] , and the series of seminal papers by Kimura [4–6] . A number of stochastic models have been used to study evolution in finite populations , of which the Moran process and the Write Fisher process are perhaps the best known . The Moran process [7] assumes the existence of N individuals , and dynamics are modeled as a sequence of updates , such that each time one individual is chosen to be removed , and another is chosen for reproduction ( thus keeping the total population size constant ) . N such elementary updates correspond to one generational update . In the Wright Fisher process ( see e . g . [8] ) , the next generation is populated by randomly drawing ( with replacement ) copies of individuals from the current population . One of the central questions that has attracted attention of researchers in the last several decades is the role of the population structure in the evolutionary dynamics . This research was pioneered by Kimura and Weiss who were the first to include spacial structure in population models [6] . Maruyama analyzed the fixation behavior of a Moran process on regular spatial structures and discovered that the fixation probability is independent of the spatial structure of the population ( for example , fixation probability on regular graphs is the same as that on unstructured graphs ) [9 , 10] . Liberman et . al extended the analysis to arbitrary graphs ( networks ) [11] . They showed that some networks may act as amplifiers , and others as suppressors of selection . Namely , amplifier graphs increase ( decrease ) the fixation probability of advantageous ( disadvantageous ) and mutants; suppressors , on the contrary , decrease ( increase ) the fixation probability for advantageous ( disadvantageous ) mutants [12–14] . In [15] , the role of the order of the update events ( birth and death ) for evolutionary dynamics in the Moran process was studied . It was discovered that for 1D and 2D spatial lattices , the fixation probabilities for birth-death and the death-birth formulations are significantly different . Apart from fixation probability , the average fixation time is an important characteristic of birth and death processes . Much research has been devoted to studying mathematical properties of this quantity in various contexts . Frean and Baxter analyzed the mean fixation time of a mutant for two homogeneous and heterogeneous graphs [16] . They have considered four different update rules of birth-death ( BD ) and death-birth ( BD ) processes for star and complete graphs: B-FD ( birth depends on fitness and death is uniform ) , B-DF ( birth is uniform and death depends on the unfitness ) , D-BF ( uniform death and fitness dependent birth ) , and DF-B ( fitness dependent death and uniform birth ) . They have shown that the star is a suppressor in both DB processes and an amplifier in both BD cases . For further developments in the studies of the evolutionary dynamics on graphs , one can refer to the review by Shakarian et al . [17] , where the authors describe the original models for evolutionary graph theory and its extensions , as well as the calculation of the fixation probability and time to fixation . Broom et al . [18] have studied the evolutionary game theory of finite structured populations with invasion process updating rules . The exact solutions are presented for the fixation probability and time for the case that mutants have fixed fitness and the case where the fitness of individuals depends on games played among the individuals , on the star , circle and complete graphs . [19] studied the importance of fixation time for the rate of evolution and showed that in star-structured populations , evolution can slow down even while selection is amplified . Hindersin and Traulsen used analytical calculations to find the fixation time of a single mutant for small graphs [20] . They showed that , interestingly , there is no obvious relation between fixation probability and time . More recently , Askari and Aghababaei-Samani introduced an exact analytical approach in order to calculate the mean time fixation of a mutant for circle and star graphs [21] . In a number of previous studies , the evolutionary properties of mutants have been investigated under the assumption that fitness values of different types were kept constant . It has been recently recognized , however , that fluctuating fitness values can have important effects on the fixation probability and time [22 , 23] . In [24] , the authors considered two different types of heterogeneity , a heterogeneous voter model where each voter has an intrinsic rate to change state and a partisan voter model where each voter has an innate and fixed preference for one opinion state ( 0 or 1 ) . Using a mean-field approximation , they compared the time to fixation for each of these two models and studied the population-size dependency of the time to fixation ( i . e . the time to ultimately reach consensus ) . Rivoire et al . [25] have used mathematical modeling and stochastic control theory to quantify phenotypic variation schemes , which are inherited , randomly produced , or environmentally induced , and have analyzed the adaptation towards such variations . Moreover , Melbinger and Vergassola [26] considered the effect of environmental alterations on the fitness of species . They showed that variability in the growth rates played an important role in neutral evolution and that the fixation time was reduced in the presence of time dependent environmental fluctuations . In addition , Cvijović et al . In general , studies of aspects of evolutionary dynamics related to heterogeneity have given rise to a number of interesting papers . In paper [27] , a different type of population heterogenity is studied: the individuals differ by the number of connections they have with others . Paper [28] studies the fixation times in both death-birth and birth-death processes under different assumptions on the underlying network structure , the game , and fitness definition . [29] found that temporal fluctuations in environmental conditions could influence the fate of mutation and subsequently the efficiency of natural selection . They have shown that temporal fluctuations can reduce the efficiency of natural selection and increase the fixation probability of mutants , even if they are strongly deleterious on average . In [22] , we used a number of models ( several versions of the Moran model and the haploid Wright-Fisher model ) to examine fixation probabilities for a constant size population , where the fitness was a random function of both allelic state and spatial position . Namely , it was assumed that the fitness values of wild type cells and mutant cells were drawn from probability distributions , and were fixed for each location . Different scenarios were examined , including correlated and uncorrelated fitness values of wild type cells and mutants , and different underlying population structures ( circles and complete graphs ) . In the case where the probability distribution of mutant and wild type cells were identical , our model of spatial heterogeneity redefined the notion of neutrality for a newly arising mutation , as such mutations fixed at a higher rate than that predicted under neutrality . In particular , it was found that the probability of mutant fixation ( in the case when the mutants were initially a minority ) was significantly larger than their initial fraction , and this effect increased with N . In other words , mutants behaved as if they were selected for , although on average their fitness values were the same as the the fitness values of wild type cells . In the current paper we investigate the question of the timing of mutant fixation in a similar setting . Does spatial randomness of this sort influence the rate of mutant fixation , and if so , in what way ? We use both analytical and numerical methods to answer this question under a variety of different assumptions on the probability distributions of wild type and mutant fitness values , and examine different biologically relevant scenarios . In particular , we investigate if spatial arrangement of cells plays a role in the timing of mutant fixation .
Suppose a population consists of two types of individuals ( or cells ) , A ( the wild type ) and B ( the mutant ) . Consider a death-birth ( DB ) formulation of the Moran process . At each update , a death event is followed by a division event , where the removed individual is replaced by the offspring of one of its neighbors . While individuals are chosen for death randomly with equal probabilities , reproduction probability of each is proportional to its division rate , or fitness . Traditionally , fitness of individuals is defined by their type , such that the wild type and mutant cells have fitness values rA and rB respectively . The notion of neighborhood is defined by a graph connecting the individuals . For example , one may consider the complete graph , where any cell could be chosen for division to replace any eliminated cell . Another example of a graph is a circle , where each cell only has two neighbors . In constant population processes with two sub-populations , in the absence of mutations , there are only two absorbing states: the state where all cells are wild type , and the state where all the cells are mutants . If starting from a nonzero number of mutants , the system reaches the all-mutant state , we say that the mutants have reached fixation ( and otherwise we say that they have gone extinct ) . The expected time of fixation conditioned on the event of mutant fixation , ti , has been calculated for several graphs . Antal and Scheuring [30] have used an evolutionary game model to calculate fixation of strategies in finite populations . Recently , Hindersin and Traulsen [20] have obtained the fixation time for all possible connected networks with four nodes ( 6 graphs ) . Unlike the traditional Moran process , here we consider the case where the fitness of individuals depends on their environment , in the following sense . In a system of N individuals , the fitness depends not only on whether an individual is wild type or mutant , but also on the location of each individual on the graph . We assume that each vertex of a graph is associated with a ( fixed ) wild type fitness parameter , which is generated randomly according to a given distribution , and does not change in time . Similarly , each spot is also characterized by a mutant fitness parameter . These parameters are also chosen randomly from a distribution , they characterize the fitness of mutants at different locations , and remain constant in time . The wild type and mutant fitness probability distributions may in general be the same or different , and the choice of mutant and wild type fitness values could be uncorrelated or correlated to different degrees [22] . Here , for illustration purposes , we will consider a discrete symmetric bimodal fitness distributions , such that fitness parameters are randomly selected to be 1 + σ or 1 − σ , with 0 < σ < 1 . In addition to the DB formulation of the Moran process described above , we also studied the birth-death ( BD ) formulation of the Moran process , and the haploid Wright Fisher model . In the BD Moran process , for each update , first a cell is selected for reproduction ( based on the cells’ fitness values ) , and then a neighboring cell ( excluding the reproducing cell itself ) is chosen to be removed , such that all candidate cells have the same probability to be chosen . The offspring of the reproducing cell replaces the removed cell . In the haploid Wright Fisher model , each new generation of cells is created by randomly sampling ( with replacement ) the cell types from the current population . The probability to be picked is proportional to the cells’ fitness . Most of the ideas of this paper are illustrated by using the DB Moran model . The results obtained from the other models are similar and are presented in S1 Text . To find the mean conditional fixation time , we use Chapman-Kolmogorov equations to find the fixation properties of a mutant . In a population of N individuals , there are M = 2N distinct states . This can be shown by observing that in the presence of m mutants , 0 ≤ m ≤ N , there are N ! /m ! ( N − m ) ! distinct configurations; summing those up gives the value 2N . For each fixed fitness realization , γ , let us denote by T i → j γ the probability of transitioning from state i to state j , 1 ≤ i , j ≤ M . Let us denote the absorbing state ( or the set of absorbing states ) of interest as E . In our particular case , E is the state where each location contains a mutant ( mutant fixation ) . The other absorbing states comprised of the set E1 ( in our case this is the state of all wild type cells , that is , mutant extinction ) . Following [30] , denote by ρ i γ ( t ) the probability to get absorbed in state E starting from state i after t steps , under fitness configuration γ . The total probability of absorption in E is then given by ρ i γ = ∑ t = 0 ∞ ρ i γ ( t ) . We have for this quantity , ρ i γ = ∑ j T i → j γ ρ j γ , i ∉ E , E 1 , ( 1 ) where the summation in j goes over all the states , and ρ E γ = 1 , ρ E 1 γ = 0 . ( 2 ) This is a linear system of M − 2 equations for ρ i γ ( the Chapman-Kolmogorov equation [31] ) . Next , let us denote by t i γ the mean conditional time it takes to get absorbed in state E starting from state i under configuration γ , given that absorption happens: t i γ = ∑ t = 0 ∞ t ρ i γ ( t ) ∑ t = 0 ∞ ρ i γ ( t ) = ∑ t = 0 ∞ t ρ i γ ( t ) ρ i γ , and further denote τ i γ = t i γ ρ i γ . We have ρ i γ ( t ) = ∑ j T i → j γ ρ j γ ( t - 1 ) . ( 3 ) Changing the summation index , we obtain ∑ t = 0 ∞ t ρ j γ ( t - 1 ) = ∑ t = 0 ∞ ( t + 1 ) ρ j γ ( t ) = τ j γ + ρ j γ . Multiplying eq ( 3 ) by t and summing up from 0 to infinity , we obtain τ i γ = ∑ j T i → j γ τ j γ + ∑ j T i → j γ ρ j γ , or equivalently , τ i γ = ∑ j T i → j γ τ j γ + ρ i γ , i ∉ E , E 1 , ( 4 ) where the summation in j goes over all the states , and τ E γ = 0 , τ E 1 γ = 0 . ( 5 ) Eqs ( 1 ) and ( 4 ) comprise a closed system that can be solved for ρ i γ and τ i γ for all i ∉ E , E1 . Then , the conditional mean time of absorption under configuration γ is given by t i γ = τ i γ ρ i γ . We are interested in the expectation of this quantity over all realizations of the fitness realization , γ: E γ [ t i γ ] = E γ [ τ i γ ρ i γ ] . ( 6 ) As an example , we apply this theory to the circle graph in the context of the death-birth Moran process . For illustration purposes , we use the population size N = 3 ( note that in this case , the circle and the complete graph are the same ) . Denote the states of the Markov chain by vector ( n1 , n2 , n3 ) , where ni = 1 if the site is occupied by a mutant and ni = 0 otherwise . There are two absorbing states: ( 000 ) , the state occupied with all wild-type cells , and ( 111 ) , the state filled entirely with mutants . The states ( 100 ) , ( 010 ) and ( 001 ) are the states of Markov chain with one mutant and ( 101 ) , ( 110 ) and ( 011 ) are the states with two mutants ( Fig 1 ) . We use the notation ( a , b , c ) to denote wild type fitness values at locations ( 1 , 2 , 3 ) ; the mutant fitness values at these locations are denoted by ( a ˜ , b ˜ , c ˜ ) . For each fitness configuration , the probability of reaching the state E ( the state of all mutant cells ) starting from state ( n1n2n3 ) is defined by ρ n 1 n 2 n 3 . Using formulas ( 1 ) and ( 2 ) , one can write Chapman-Kolmogorov equations for the fixation probability under a fixed set of fitness values , 3 ρ 100 = a ˜ a ˜ + b ρ 101 + a ˜ a ˜ + c ρ 110 + ( b a ˜ + b + c a ˜ + c ) ρ 100 , 3 ρ 010 = b ˜ b ˜ + c ρ 110 + b ˜ b ˜ + a ρ 011 + ( c b ˜ + c + a b ˜ + a ) ρ 010 , 3 ρ 001 = c ˜ c ˜ + a ρ 011 + c ˜ c ˜ + b ρ 101 + ( a c ˜ + a + b c ˜ + b ) ρ 001 , 3 ρ 101 = b a ˜ + b ρ 100 + b c ˜ + b ρ 001 + 1 + ( a ˜ a ˜ + b + c ˜ c ˜ + b ) ρ 101 , 3 ρ 110 = c a ˜ + c ρ 100 + c b ˜ + c ρ 010 + 1 + ( a ˜ a ˜ + c + b ˜ b ˜ + c ) ρ 110 , 3 ρ 011 = a b ˜ + a ρ 010 + a c ˜ + a ρ 001 + 1 + ( b ˜ b ˜ + a + c ˜ c ˜ + a ) ρ 011 . ( 7 ) Denote by ρ˜m the fixation probability starting with m mutants , averaged over all realizations of the fitness value sets . For N = 3 , if the mutant and wild type fitness values are generated from the same distribution , the average fixation probability is a constant independent of the probability distribution [22]: ρ˜ 1= 1 3 , ρ˜2 = 2 3 , ( 8 ) which coincides with the result ρm = m/N for neutral mutants , in the absence of randomness . Note that for larger values of N ( namely , all N > 3 ) , the mutant fixation probability is no longer m/N , and it depends on the distribution . In particular , for minority mutants ( that is , for m < N/2 ) , the probability of fixation increases with the variance of the underlying distribution [22] . Denoting by t n 1 n 2 n 3 = τ n 1 n 2 n 3 / ρ n 1 n 2 n 3 the mean fixation time needed for going from state ( n1n2n3 ) to state ( 111 ) , under a fixed fitness realization , we obtain the following six linear Chapman-Kolmogorov equations ( see eqs ( 4 ) and ( 5 ) ) : 3 ρ 100 = - a ˜ a ˜ + b τ 101 - a ˜ a ˜ + c τ 110 + ( 1 + a ˜ a ˜ + b + a ˜ a ˜ + c ) τ 100 , 3 ρ 010 = - b ˜ b ˜ + c τ 110 - b ˜ b ˜ + a τ 011 + ( 1 + b ˜ b ˜ + c + b ˜ b ˜ + a ) τ 010 , 3 ρ 001 = - c ˜ c ˜ + a τ 011 - c ˜ c ˜ + b τ 101 + ( 1 + c ˜ c ˜ + a + c ˜ c ˜ + b ) τ 001 , 3 ρ 101 = - b a ˜ + b τ 100 - b c ˜ + b τ 001 + ( 1 + b a ˜ + b + b c ˜ + b ) τ 101 , 3 ρ 110 = - c a ˜ + c τ 100 - c b ˜ + c τ 010 + ( 1 + c a ˜ + c + c b ˜ + c ) τ 110 , 3 ρ 011 = - a b ˜ + a τ 010 - a c ˜ + a τ 001 + ( 1 + a b ˜ + a + a c ˜ + a ) τ 011 . ( 9 ) The mean conditional fixation time averaged over all realizations of the fitness landscapes is then obtained from eq ( 6 ) . For N = 3 , we can solve the above equations analytically , and then the mean conditional fixation time ( averaged over all realizations of the fitness values ) is obtained from eq ( 6 ) . We have written a Mathematica code that generates the set of equations for any N ( the equations for N = 4 are presented in S1 Text ) . However , this approach is only practical for small values of N ( due to the large number of equations and configurations ) . For larger networks , we use the canonical matrix method [32] ( see S1 Text ) or stochastic simulations to find the fixation probability and the mean conditional fixation time . The equations described above only have practical applicability for relatively small networks . As N increases , the number of equations grows as 2N − 2 for the complete graph and N ( N − 1 ) for the circle . Instead of solving the algebraic equations , stochastic numerical simulations have to be implemented . In the numerical simulations , we consider the population on a graph , where each vertex is an individual ( wild type or mutant ) . For each simulation , we generate wild type fitness values and mutant fitness values from their respective probability distributions . These values are associated with their vertices on the graph and are kept constant until the end of the simulation . Since the fitness values are randomly selected from a bimodal distribution ( 1 + σ and 1 − σ ) for each individual at every node of the graph , there will be 22N fitness configurations . We start with an initial condition of one mutant ( type B ) and N − 1 wild types ( type A ) . At each time step , as long as the mutant population has not yet become fixated or gone extinct , one individual is randomly removed and one of its neighbors is chosen for division with a probability proportional to its fitnesses . The simulation is stopped when the mutants become extinct or reach fixation . If mutant fixation is reached , we record the number of updates until fixation; this gives the time to fixation for a particular ( successful ) run . This process is repeated a number of times for each configuration . The mean conditional time for each configuration is the sum of all individual fixation times divided by the number of successful samples ( that is , the number of runs where mutant fixation was observed ) . The overall mean conditional fixation time is the average of the mean conditional fixation times over all possible configurations . A computational difficulty with this approach arises from the fact that for some fitness configurations , the probability of mutant fixation is very low . For such configurations , after running the simulation a fixed number of times , it may happen that fixation never occurs , in which case the configuration will not contribute into the calculated mean conditional fixation time . The configurations with low fixation probabilities are less likely to be fixed , which may skew the numerical results . In order to avoid this problem , we executed over 106 independent realizations for each configuration . As a result , the simulation is very costly . For larger networks , instead of the exhaustive calculation described above , we used a sampling method similar to the method that was implemented in [22] . Since listing all the configurations becomes computationally impossible , we only looked at a subset of possible realizations of random fitness values . For each such realization , we ran simulations starting from one mutant cell , until the mutants reached fixation . At this point , the time it took to fixate was recorded , and we moved on to the next randomly chosen configuration . The mean conditional fixation time was then approximated as an average over fixation times obtained for these realizations .
We use Chapman-Kolmogorov equations presented above to calculate the mean conditional fixation time . We start by examining the case of circular networks . So far , all the calculations were performed for small circular networks . Next , we turn to complete graphs . In Fig 4 we perform calculations on a complete graph for N = 4 through N = 6 ( panel ( a ) - ( c ) ) and show the dependence of mean conditional time on σ . In contrast with the results for the circular graphs , the type of dependence does not change with system size , N . As expected , for σ = 0 the non-random values 9 , 16 , 25 are obtained . As the value of σ increases , the mean time to fixation decreases . As will be shown below , for larger values of N , the mean conditional time is also a decreasing function of σ . This result is quite general . In S1 Text , we extended our calculations for the mean conditional fixation time on complete graphs to the BD formulation of the Moran process and to the ( haploid ) Wright Fisher model [22] . In these cases , the mean conditional fixation time is also a decreasing function of σ . The effect of correlations between mutant and wild-type fitness values is studied in Fig 4 ( a ) –4 ( c ) . As for circular graphs , the mean conditional fixation time is always the largest for the fully correlated case and the smallest for the anti-correlated case . Fig 4 ( d ) shows the dependence of fixation time on skewness in the case of N = 6 ( other values of N show similar trends ) ; the behavior is again qualitatively similar to that of the circular graphs . Finally , in S1 Text we studied the probability distribution of the time to fixation for nonzero σ . As in the case of small circles , the distribution of fixation times becomes wider with N . So far we have considered the scenario where the probability distributions of the wild type and mutant fitness values were the same . Here we allow them to be different , and study two interesting cases: ( a , b ) only one of the two types has random fitness values , while the other is deterministic , keeping the same mean fitness; ( c , d ) both types are random , but one of the types is advantageous . Fig 5 explores the above cases for the N = 3 network . We can see large differences in the behavior of fixation times , depending on which type is random and which type is advantageous . For case ( a ) where only the fitness of mutant individuals is changing and case ( d ) where the wild-types are advantageous , the fixation time is a decreasing function of the standard deviation σ . For case ( b ) , where the mutant cells are assumed to have fixed fitness 1 and the wild-types have a random fitness with average 1 , and case ( c ) where the mutants are advantageous , the fixation time is an increasing function of standard deviation . Interestingly , these results do not generalize for larger values on N , and the complexity disappears as N increases . What we find is that all the four situations exhibit similar dependencies , and the main factor determining the behavior is the type of network . In Fig 6 , we increase the population size to N = 6 and N = 7 and study cases where only one of the species ( mutants or wild types ) have random fitness , whereas the other species has a fixed fitness value , with both fitness values having the same mean . Panels ( c ) and ( d ) correspond to complete graphs , and it is clear that randomness accelerates fixation . Panels ( a ) and ( b ) show the results for circular graphs . While some non-monotonicity is present for the case of random wild types and deterministic mutants , it disappears for larger N , and the general result for circles is that randomness delays fixation . Fig 7 studies the case where either wild types or mutants are advantageous , while both cell types have random fitness . Again , randomness delays fixation for circular graphs ( a ) and accelerates it for complete graphs ( b ) . These results are consistent with the rest of the findings for circular networks and complete graphs . We delay an intuitive explanation for this phenomenon until the next section . To investigate the effect of random environment on the mean fixation time for larger networks , we turn to stochastic simulations . Again , we consider the complete graph and circle arrangement with different population sizes . The simulated results are given for the DB Moran process in the case where the fitnesses of both kinds of individuals are selected from random ( binomial ) distribution with average one , i . e . 1 + σ or 1 − σ . First , we investigate the impact of random fitness on the mean conditional fixation time of a mutant for circle and complete graph with different population sizes ( Fig 8 ) . We observe that , as expected , the larger the population size N , the larger the fixation time of the mutants . Further , for equal values of N , the fixation happens faster on a complete graph than on a circle . This is also expected , as there are more pathways for mutants to spread on a complete graph , compared to a circle where a ( one-dimensional ) mutant patch can only grow through its two boundaries . Next , we explore the dependence of the mean fixation time on randomness . Recall that in the case of circles ( see Fig 2 ) , for N = 3 , the mean conditional fixation time increased with σ , it decreased with σ for N = 4 , was non-monotonic for N = 5 , and increased again with N = 6 . It turns out that the trend observed for N = 6 persists for larger values on N , see Fig 8 ( a ) , where we can see that the mean conditional fixation time increases with standard deviation . Interestingly , the result for the complete graph is very different ( Fig 8 ( b ) ) . There , the mean conditional fixation time is a decreasing function of the standard deviation for all population sizes . The explanation for this phenomenon is quite intuitive . As mentioned above , on a circle , the mutant population spreads out from a single mutant as a connected patch . This patch must expand to the whole circle to reach fixation , and the presence of a fitness “dead zone” ( a sequence of several consecutive low fitness values for the mutants on the random fitness landscape ) serves as a hurdle that can significantly increase fixation time , as there is no way around those dead zones . On the other hand , a complete graph allows many “paths” to fixation , because every spot is everyone’s neighbor , and the presence of several low fitness spots does not preclude the mutants from spreading in the same way as it does in a 1D geometry . Moreover , for a fully connected graph , the presence of randomness actually creates opportunities , increasing the likelihood of “lucky” paths to fixation , where several “neighboring” spots have an elevated mutant fitness . This explains a decrease in the expected fixation time as randomness on a complete graph increases , Fig 8 ( b ) . We note that for small circles , the dependence of the mean conditional fixation time on randomness is less straightforward , because for very small networks the difference between the number of pathways to fixation on a circle and on a complete graph is not as drastic as it is for larger N . This explanation of the effect of randomness on fixation time holds also for the scenarios where the fitness probability distributions are different for mutants and wild type cells . The following scenarios of interest were studied in the previous section . ( i ) The expected fitness value is the same for the wild types and the mutants , but either the wild type or mutant fitness values are constant ( non-random and equal to their expectation ) , while the other type’s fitness values have a nonzero variance . ( ii ) Both types have random fitness values , but the mean fitness of mutants is larger or smaller than that of the wild types . In all these cases , it was observed that for sufficiently large values of N , the mean conditional mutant fixation time is an increasing function of randomness for circles and a decreasing function of randomness for complete graphs . To understand the drastic changes in the behavior for small networks ( Fig 5 ) and larger networks ( Figs 6 and 7 ) , let us first turn to Figs 6 ( b ) and 7 ( a ) , which describe larger circles and contain all 4 cases ( a , b , c , d ) listed in Fig 5 . They all show an increase in the fixation time as randomness increases , which coincides with our prediction for circles . Next , consider Figs 6 ( a ) and 7 ( b ) , both of which describe larger complete graph networks . Again , these two figures contain all 4 cases ( a , b , c , d ) , and they all show a decrease in the fixation time as randomness increases . This is the result that we predict for complete graphs . The way one could intuitively understand the behavior of the N = 3 network of Fig 5 is to remember that this network is simultaneously a circle and a complete graph . It turns out that when it comes to deterministic or advantageous mutants , the N = 3 network behaves as if it was a circle . In the case of deterministic or advantageous wild types , it behaves as if it were a complete graph . For larger networks this “confusion” disappears and the results are consistent with our intuition . The ideas presented here are illustrated further when we compare the results of mean fixation time for regular graphs with different degrees ( see Fig 9 ) . Define the variable z as the half of the number of neighbors for each node . For a regular graph with size N = 9 , the parameter z = 1 corresponds to the circle structure with the nearest neighborhood for each node , z = 2 corresponds to the circle with the second nearest neighborhood and so on , until z = 4 which corresponds to the complete graph . Increasing the value of z , the mean fixation time decreases , and as a function of σ , it also switches from an increasing to a decreasing function . We have also studied the behavior of mutant fixation behavior as a function of the initial number of mutants . We have calculated the unconditional absorption time , which is the expected time to get into either of the two absorbing states ( all mutants or all wild-type cells ) . The results are presented in S1 Text and are consistent with the rest of the findings: unconditional mean absorption time grows with randomness for circles and decreases for complete graphs .
In summary , we have studied several constant population models ( two formulations of the Moran model and the Wright Fisher model ) , where the fitness values of cells depend not only on their types ( mutant or wild type ) but also on their spatial locations , representing environmental factors . Fitness values of mutants and wild type cells at different locations are drawn from fixed probability distributions and remain constant in time . We ask how mutant ( conditional ) fixation times are influenced by this type of environmental randomness . Before we summarize the results , we want to emphasize the applications that motivated this study and to show that our approach is rooted in real biological problems . Let us think of a spatially distributed population , say , a number of plants in an expanse of soil . It would be quite natural to assume that some spots can be more favorable and others less favorable . Examples of factors that contribute to fitness are sunlight , proximity of water , soil quality , the presence of rocks etc . Now , let us suppose that a mutant reacts differently to the same variations in the environment . Assume that a plant species does not well tolerate the presence of rocks in the soil , and that a mutant plant is more tolerant to the presence of rocks , but is very sensitive to the sunlight . Then the fitness values of the two subspecies on a spatial grid will be different , and defined by the location and by the mutation status , as assumed in our model . In the extreme case , wild type fitness is only defined by the absence of rocks , and mutant fitness is only defined by sunshine . In this case , the two fitness value sets are uncorrelated . If both factors play a role , but to different degrees , in the plants’ fitness values , then the fitness values will exhibit a degree of correlation , as described in this study . Similar considerations apply to a large variety of biological settings . The effects of spatial structure and heterogeneity are important in biological models such as bacterial growth , where fitness can be a function of the spatial distribution of nutrients and microenvironment . In [33] it was demonstrated clearly that in biofilms , there are significant spatial microscale heterogeneities , both in chemical and physical parameters of the biofilm interstitial fluid . For example , complex patterns of water flow with different velocities and directions were observed throughout the biofilms . Further , heterogeneity in solute chemistry that is present within a biofilm was reported including concentration gradients of metabolic substrates and products . It was also reported that microorganisms within the biofilms can and do respond to these local environmental conditions in a variety ways , such as altering gene-expression patterns or physiological activities . Mutations arising and spreading in bacterial populations lead to high levels of genotypic and phenotypic heterogeneity in biofilms . It has been proposed that such diversification of bacterial populations may be considered an adaptation to the microscopic scale heterogeneity of the environment [34 , 35] , as different phenotypes respond differently to the changes in the environment . Diverse populations have been described as more robust; the “insurance hypothesis” states that the presence of diverse subpopulations increases the range of conditions in which the community as a whole can survive . From the theoretical prospective it is therefore essential to understand the evolutionary dynamics of mutations in the environment characterized by microscale heterogeneities . Another important application of our theory is dynamics of cancer cell populations . It is well known that solid cancers are characterized by a highly complex and heterogeneous microenvironment [36] , which includes stroma , necrotic cells , blood vessels , etc . The distribution of oxygen and hypoxic regions is highly non-homogeneous [37] , the nutrients are distributed in a complex fashion , and in general , no two tumors are the same [38] . Tumours have been compared to unhealed wounds [39] , in that they produce large amounts of inflammatory mediators ( cytokines , chemokines , and growth factors ) . These molecules attract the so called tumor infiltrating cells that include macrophages , myeloid-derived suppressor cells , mesenchymal stromal cells , and TIE2-expressing monocytes . Together , these populations of non-imalignant cells contribute to the formation of a rich and heterogeneous tumor microenvironment [40] . In order to understand selection and mutation dynamics of cancer cell populations in such an environment , it is not enough to restrict the modeling efforts to the classic problems , where all the wild type cells are exactly ( phenotypically ) the same and all the mutant cells have the same constant fitness value . In this study , we make a step towards a more realistic view of cancer dynamics , where fitness values of different genotypes are subject to microenvironmental variations . In our study , we consider several different scenarios , where we vary assumptions on the probability distributions underlying the mutant and wild type fitness values . In particular , we investigate the cases where All scenarios are investigated in the context of two types of networks: circles and complete graphs . We find that the results are very different for these two choices of the underlying network . It turns out that environmental randomness has a significant effect on the conditional fixation time of mutants . A clear trend was observed when studying the behavior of mutants on the two different networks: randomness delayed fixation of mutants on circles ( at least for values of N larger than a threshold ) , and it accelerated fixation on complete graphs . The reason is that for 1D—type structures ( circles ) , “dead zones” that form randomly in the presence of environmental influences , can significantly delay fixation by blocking the paths to fixation . For fully connected graphs , “lucky paths” form at random , that facilitate fixation . These trends have been observed for all the scenarios above , except for very small circular networks ( N ≤ 5 ) some additional complexity was present . Otherwise , this pattern was universal and included scenarios with identical and different fitness probability distributions for mutants and wild type cells , in the presence of a deterministic type , and in the presence of advantageous/disadvantageous types . Next , when studying the effects of correlations , we observed that in the case of both circles and complete graphs , the mean conditional fixation time was the largest in the fully correlated case and the smallest in the anti-correlated case . Finally , if the probability distribution underlying the fitness realizations is skewed , large negative skewness values increase the effect of randomness on the mean conditional fixation time . This is observed in all the scenarios , whether the effect was accelerating or decelerating . Large negative skewness implies the existence of rare but very disadvantageous spots . These spots can serve as a serious impediment in the constrained circular geometry . In complete graphs , they do not present a problem because of the existence of multiple paths to fixation; at the same time having most spots with a slightly elevated fitness facilitates fixation . The difference between fixation probabilities in circular and complete graphs exemplifies the general phenomenon of well-mixedness and its role in evolutionary mutant dynamics; it is further related to the role of dimensionality in system dynamics . Circular graphs are one-dimensional systems , where the geometric or spatial constraints are the most rigid . The opposite scenario is presented by the complete graphs , which correspond to the mass-action or complete mixing assumption . The spatially arranged two- and three-dimensional grids are somewhere between these two extreme scenarios , with the three-dimensional arrangement being closer to mass-action . We have studied evolutionary dynamics in these different settings in several contexts . For example , it has been shown that inactivation of a tumor-suppressor gene ( a two-hit evolutionary process in which the cells must first become less fit before becoming more fit ) happens faster in 1D ( a row of cells ) [41 , 42] , than in 2D ( a layer ) , and this is in turn faster than in a fully mixed system with no spatial constraints [41 , 43–45] . By contrast , in two-step processes in which the intermediate mutant confers a slight selective advantage , the relationship is the opposite , and a non-spatial , fully mixed environment promotes the fastest pace of evolution [45] . As was commented in [46] , these phenomena seem less surprising if one notes how reminiscent they are of other fundamental laws of nature in which space dimensionality changes how things work , such as the different fundamental solutions of Poisson’s equations in 1D and 2D . We have presented results for both small N and large populations . We anticipate that there are interesting applications even for small N . For instance , an important biological application of the ring geometry is the model of a human colonic crypt , where the relatively small ( of the order of ten cells or less ) population of the stem cells is situated along circular bands [47] , which can be viewed as cross-sections of three-dimensional crypts . In this context , fixation is referred to as monoclonal conversion . Although the details of the exact composition of the colonic crypt is still being debated , many researchers believe that the active stem cells occupy a narrow layer , and divide mostly symmetrically . The two division types , proliferation and differentiation , are mathematically equivalent to divisions and deaths in our models . The origins of colon cancer can be studied by examining selection dynamics of mutants in such a system . Very interesting and non-trivial is the connection between fixation time and fixation probability . These two measures of mutant success may be positively or negatively correlated for different graph structures . In particular , it appears that for a circle , randomness makes fixation longer , but it also makes it more likely; for the complete graph , randomness makes fixation both faster and more likely [22] . Counterintuitive properties of the fixation time in network structured populations were already noticed in [20] . This paper suggested that there was no obvious relation between the fixation probability and fixation time for a mutant on a network . Based on small networks , the authors analytically showed that: ( i ) Although the fixation probability was the same for all regular graphs ( for example , circle and diamond ) , it would take different times for a single mutant to get fixated on these kinds of networks . ( ii ) The graphs that were amplifiers of selections ( for example , star or line ) , increased the fixation probability of the mutant , but at the same time they could slow down the fixation process .
|
We study the influence of randomness on evolutionary dynamics , assuming that a newly arising mutant may experience a different set of environments compared to the wild type . We calculate the mean conditional fixation time of the mutant under different assumptions on spatial interactions , and show that randomness has a strong impact on the fixation time . In particular , it delays the fixation of mutants on 1D circles and accelerates it on complete graphs ( the so called mass action , or complete mixing , model ) . This result holds for advantageous , disadvantageous , and neutral ( on average ) mutants . The reason for this pattern is quite intuitive: in a rigid , 1D structure , randomness can by chance put a “roadblock” and disrupt mutant spread , causing significant delay . In higher dimensions , there are many ways for a mutant to spread , and it is difficult to block all of them by chance; on the other hand , randomness can enhance fixation by providing an “easier” path . The effects of a random environment are important in biological models such as bacterial growth or cancer initiation/progression .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"bacteriology",
"infographics",
"biofilms",
"population",
"dynamics",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"microbiology",
"probability",
"distribution",
"mathematics",
"population",
"biology",
"skewness",
"computer",
"and",
"information",
"sciences",
"probability",
"theory",
"population",
"metrics",
"bacterial",
"biofilms",
"population",
"size",
"data",
"visualization",
"cell",
"biology",
"natural",
"selection",
"graphs",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"evolutionary",
"biology",
"evolutionary",
"processes"
] |
2017
|
The effect of spatial randomness on the average fixation time of mutants
|
Aspergillus fumigatus causes invasive pulmonary disease in immunocompromised hosts and allergic asthma in atopic individuals . We studied the contribution of lung eosinophils to these fungal diseases . By in vivo intracellular cytokine staining and confocal microscopy , we observed that eosinophils act as local sources of IL-23 and IL-17 . Remarkably , mice lacking eosinophils had a >95% reduction in the percentage of lung IL-23p19+ cells as well as markedly reduced IL-23 heterodimer in lung lavage fluid . Eosinophils killed A . fumigatus conidia in vivo . Eosinopenic mice had higher mortality rates , decreased recruitment of inflammatory monocytes , and decreased expansion of lung macrophages after challenge with conidia . All of these functions underscore a potential protective role for eosinophils in acute aspergillosis . Given the postulated role for IL-17 in asthma pathogenesis , we assessed whether eosinophils could act as sources of IL-23 and IL-17 in models where mice were sensitized to either A . fumigatus antigens or ovalbumin ( OVA ) . We found IL-23p19+ IL-17AF+ eosinophils in both allergic models . Moreover , close to 95% of IL-23p19+ cells and >90% of IL-17AF+ cells were identified as eosinophils . These data establish a new paradigm in acute and allergic aspergillosis whereby eosinophils act not only as effector cells but also as immunomodulatory cells driving the IL-23/IL-17 axis and contributing to inflammatory cell recruitment .
Aspergillus fumigatus is an opportunistic mold that produces conidia that are both small ( 2 . 5–3 μm in diameter ) and readily airborne [1] . These characteristics make A . fumigatus conidia easily dispersible , while also promoting access to the alveolar spaces in the human airway [2] . It is estimated that on average , individuals inhale hundreds of conidia a day [3] . Despite such frequent exposure , in immunocompetent hosts A . fumigatus is rarely pathogenic . Its ability to cause disease is dependent on the immunological status of the host . Thus , in immunocompromised individuals , particularly those with quantitative or qualitative phagocyte defects; conidia can germinate and invade lung parenchyma resulting in a highly lethal infection known as invasive aspergillosis ( IA ) . It is estimated that >200 , 000 people develop life-threatening IA annually [4] . In atopic patients and about 2–15% of patients with cystic fibrosis , sensitization to A . fumigatus can lead to allergic reactions that can drive asthma pathogenesis and lead to allergic bronchopulmonary aspergillosis ( ABPA ) [2 , 5] . Globally , approximately 5 million people suffer from ABPA [6] . Although eosinophilia is a hallmark of several allergic diseases including ABPA and severe asthma with fungal sensitization ( SAFS ) [7 , 8] , comparatively less is known about the involvement of eosinophils in acute aspergillosis . O’Dea et al . correlated levels of fungal cell wall chitin with eosinophil recruitment to the lungs in response to repeated aspiration of A . fumigatus conidia [9] . Lilly et al . [10] have shown that ΔdblGATA-1 mice ( which lack eosinophils ) infected with the ATCC 13073 strain of A . fumigatus conidia suffer from higher fungal burdens than WT mice . These studies also linked eosinopenia to lower levels of IL-17A two days post-infection [10] . IL-17A exists as a disulfide-linked homodimer and binds with high affinity to the IL-17RA/RC complex . IL-17A also forms a disulfide-linked heterodimer with IL-17F [11 , 12] . IL-17F can exist as a homodimer as well , binding the same IL-17R complex . For clarity , the IL-17RA/RC ligands will be referred heretofore as IL-17 , unless otherwise specified . IL-17 production is either induced or augmented by IL-23 , which is a heterodimeric cytokine composed of IL-23p19 and IL-12p40 subunits . The relationship between IL-23 and IL-17 is known as the IL-23/IL-17 axis [11] . IL-23 is among a group of cytokines that activate signal transducer and activator of transcription ( STAT ) -3 [13] . A . fumigatus readily elicits IL-23 and IL-17 production from the lungs after exposure [14] . Although recognized primarily for the induction of neutrophilia , pro-inflammatory cytokines such as IL-6 and IL-1β , and the up-regulation of antimicrobial peptides [15] , IL-17 has been reported to induce the recruitment of eosinophils in a model of chronic aspergillosis [16] . High levels of IL-17 have also been correlated with symptom severity in allergic asthma [17] . A remarkably large number of innate and adaptive immune cell types has been reported to be capable of producing IL-17 , including γδ T cells , invariant natural killer T cells , type 3 innate lymphoid cells ( ILC3s ) , neutrophils , macrophages , CD8+ T cells ( Tc17 ) and CD4+ T cells ( TH17 ) [12 , 15 , 18 , 19 , 20] . The cellular source of IL-23 has primarily been studied in connection to TH17 development . In the context of this paradigm , antigen presenting cells such as dendritic cells and macrophages have been identified as its main sources [11] . Here , we show that eosinophils are a local source of IL-23 and IL-17 in response to both acute A . fumigatus infection and in two different asthma models . In addition , we show that eosinophils are able to associate with and kill A . fumigatus conidia . We also investigate their function as immunomodulators in acute aspergillosis by regulating the recruitment of inflammatory monocytes and the expansion of macrophages in the lungs . Finally , we describe their ability to confer protection against mortality in acute aspergillosis . Taken together , our findings support a paradigm whereby eosinophils and eosinophilic production of IL-23 and IL-17 are beneficial to defenses against invasive aspergillosis but deleterious to the host in allergic disease .
A . fumigatus has been shown to stimulate the IL-23/IL-17 axis before adaptive responses are mounted [14] . Therefore , we sought to identify innate sources of the components of the IL-23/IL-17 axis in the lungs after challenge with A . fumigatus conidia . By in vivo intracellular cytokine staining ( ICS ) we uncovered a cell population co-producing IL-23p19 and IL-17A in the lungs of C57Bl/6 mice within the first eight hours of infection with 5x107 A . fumigatus conidia ( Fig 1A ) . To identify the cell type ( s ) responsible for their production , the IL-23p19+ IL-17A+ population was sorted 8 h post-infection ( Fig 1A ) . In accordance with eosinophil morphology , ≈95% of the sorted cells stained strongly with eosin in their cytoplasm , and their nuclei were characteristically polymorphous [21] . We also assessed the IL-23p19+ IL-17A+ population in BALB/c mice acutely infected with A . fumigatus and found the cells were Siglec-F+ CD11b+ and CD11c- ( Fig 1B ) . This pattern of surface marker expression has been shown to accurately define eosinophils in the murine lungs [22] . However , in mice where eosinophilopoiesis is disrupted ( ΔdblGATA-1; BALB/c background ) , IL-23p19+ IL-17A+ cells were largely absent ( Fig 1B ) . To further confirm eosinophils express IL-17A , we infected IL-17AGFP/GFP mice in the C57Bl/6 background . Due to the autofluorescence exhibited by eosinophils and alveolar macrophages in the channel used to detect GFP signal [23 , 24] , we stained cells with a monoclonal antibody against GFP conjugated to Alexa Fluor 647 . We found that eosinophils ( Siglec-F+ CD11c- ) indeed expressed GFP in the IL-17A reporter mouse line ( Fig 1C ) . IL-23 and IL-17 production is not a universal feature of all fungal stimulated eosinophil populations as co-incubation of bone marrow-derived eosinophils ( BM-eos ) with A . fumigatus conidia , zymosan or LPS elicited levels of these cytokines that were below the lower limits of detection as measured by an ELISA that was sensitive to 4 pg/ml . In addition , there was no signal over baseline when the stimulated BM-eos were examined by ICS . Levels of IL-23 and IL-17 remained undetectable even following stimulation of the bone marrow-derived eosinophils with IL-5 , IL-1β , GM-CSF , IL-17E , PGE2 , transforming growth factor-β and IL-6 over the time and concentration ranges stated in Materials and Methods . Moreover , stimulation with IL-23 failed to elicit detectable IL-17 ( <4 pg/ml ) and stimulation with IL-17AA did not result in detectable IL-23 ( <4 pg/ml ) . To our knowledge , this is the first report showing that eosinophils express IL-23p19 . To further explore this finding , we looked for pulmonary IL-23 sources by comparing eosinophil-deficient ΔdblGATA-1 and wild-type BALB/c mice challenged with A . fumigatus conidia ( Fig 2A ) . In comparison to BALB/c mice , ΔdblGATA-1 mice exhibited a >95% reduction in the percentage of IL-23p19+ cells detected in the lungs ( Fig 2B ) . Nearly all ( 96 . 9% ) IL-23p19+ cells in WT mice were found to be in the eosinophil ( Siglec-F+ CD11c- ) gate ( Fig 2C ) . Moreover , the majority ( 87 . 1% ) of eosinophils recruited to the lungs were IL-23p19+ ( Fig 2C ) . As functional IL-23 is comprised of IL-23p19 and IL-12p40 subunits , we confirmed that IL-23 heterodimer was indeed produced in acutely infected lungs , and that their levels as measured by ELISA were significantly diminished in the absence of eosinophils ( Fig 2D ) . In conclusion , eosinophils are the predominant local source of IL-23 in this acute aspergillosis model . Given that IL-23 can induce and augment IL-17 production , we next looked at whether the low IL-23 levels found in eosinophil-deficient ΔdblGATA-1 resulted in reduced IL-17 production by inflammatory cells recruited to the lungs following A . fumigatus challenge . However , expression of IL-17AF by macrophages , inflammatory monocytes , neutrophils and non-myeloid ( CD11b- , CD11c- , Ly6G- ) cells was not significantly different comparing lung cells from wild-type and ΔdblGATA-1 mice ( S1 Fig ) . To confirm that eosinophils make IL-23p19 and IL-17A and to assess the intracellular location of these cytokines , we performed confocal microscopy on cells obtained by bronchoalveolar lavage 8 and 54 hours post-infection with A . fumigatus ( Fig 3 ) . Cells were identified as eosinophils if they had granules that stained positive for eosinophil peroxidase ( EPX , Fig 3A and 3B ) or major basic protein ( MBP , Fig 3C and 3D ) . Staining for IL-23p19 and IL-17A was observed in 97 . 7 ± 1 . 2% ( n = 130 eosinophils counted on 7 cytospin slides ) and 90 . 9 ± 2 . 6% ( n = 66 eosinophils counted on 5 cytospin slides ) , respectively , of the eosinophils examined . Moreover , IL-23p19 staining was observed exclusively in eosinophils ( Fig 3A ) . Zero out of 711 cells that did not stain for EPX or MBP examined by confocal microscopy exhibited positive staining for IL-23p19 . In contrast , IL-17A staining was more promiscuous ( Fig 3B ) . Neither cytokine co-localized on a consistent basis with MBP- or EPX-positive granules . To further characterize the innate eosinophil response to acute pulmonary aspergillosis , we assessed their recruitment over the first three days of infection . From day 1 to day 3 post-infection , the percentage of total leukocytes in the lungs that were eosinophils ( Siglec-F+ CD11c- ) did not significantly change ( Fig 4A ) . However , an increase in the number of eosinophils was observed ( Fig 4B ) . Consistent with data from Mesnil et al . [25] , uninfected BALB/c mice had detectable but relatively low numbers of resident eosinophils ( Fig 4B ) . To quantitatively assess the ability of eosinophils to associate with and kill A . fumigatus conidia in vivo , we used the Fluorescent Aspergillus Reporter ( FLARE ) assay developed by Jhingran et al . [26] . As this assay does not distinguish between conidia that are internalized from those that are surface-bound , the term “association” is used here to describe binding with or without phagocytosis . The FLARE assay exploits the instability of fluorescent proteins ( i . e . , dsRed ) in denaturing environments such as phagolysosomes to assess the fate ( i . e . , viability ) of conidia associated with a leukocyte of interest . Labeling dsRed-expressing conidia with a trace fluorophore , here we used Alexa Fluor 633 , allows for the detection of live conidia ( dsRed+ AF 633+ ) and dead conidia ( dsRed- AF633+ ) . Using the FLARE assay , the capacity of eosinophils to associate with and kill conidia in the lungs of infected BALB/c mice was assessed on days one and three post-infection . Association was calculated by adding the proportion of cells associated with live conidia ( Gate R1 ) to the proportion of cells associated with dead conidia ( Gate R2 ) . The number of eosinophils found to be associated with conidia remained constant from day one to day three post-infection ( Fig 4C and 4D ) . However , a significant increase in conidial killing was observed from day one ( 36 . 9% ) to day three ( 68 . 6% ) post-infection ( Fig 4C and 4D ) . We next considered whether eosinophils also could contribute to host defenses against aspergillosis by directly or indirectly recruiting other phagocytes to the site of infection . Therefore , we assessed the number of leukocytes ( CD45+ cells ) three days after infection with 5x107 A . fumigatus conidia in the lungs of BALB/c and ΔdblGATA-1 mice . We found that BALB/c mice had approximately 2x106 more CD45+ cells than ΔdblGATA-1 mice ( Fig 5A ) . This difference could not be accounted for by the lack of eosinophils in ΔdblGATA-1 mice , as their numbers averaged 1 . 7x105 ± 0 . 2x105 in BALB/c mice at the same 3 day time point ( Fig 4B ) . Considering that eosinophils were found to be sources of IL-23 and IL-17 in acute aspergillosis ( Figs 1 and 2 ) and as these cytokines play a significant role in the recruitment of neutrophils [27] , we evaluated the degree of lung neutrophilia one and three days post-infection in BALB/c and ΔdblGATA-1 mice . We found no differences in neutrophil numbers between groups ( Fig 5B ) . However , when we assessed the number of inflammatory monocytes ( Siglec-F- CD11b+ CCR-2+ ) and lung macrophages ( Siglec-F+ CD11c+ ) , we found that both phagocyte populations increased from day one to day three in BALB/c mice but not in ΔdblGATA-1 mice ( Fig 5C and 5D ) . In fact , at three days post-infection , ΔdblGATA-1 mice had fewer inflammatory monocytes and macrophages than BALB/c mice . These data suggest that lack of eosinophils disrupts inflammatory monocyte recruitment , which then hampers lung macrophage expansion as inflammatory monocytes differentiate into macrophages and DCs at sites of inflammation [28 , 29] . Nevertheless , percent killing of conidia by inflammatory monocytes and neutrophils was not significantly different when comparing BALB/c with ΔdblGATA-1 mice ( S2 Fig ) . Infecting BALB/c and ΔdblGATA-1 mice with 5x107 A . fumigatus conidia from the AF293 strain rendered no casualties ( Fig 6A ) . However , after infection with the same inoculum of the CEA10 strain , which has been shown to be a more virulent strain in mouse models of acute aspergillosis [30 , 31 , 32 , 33 , 34] , ΔdblGATA-1 mice had significantly higher mortality compared with BALB/c mice ( Fig 6B ) . To examine why lack of eosinophils was associated with increased mortality , we looked at CFUs in lungs , kidneys , liver and spleen 2d post infection . ΔdblGATA-1 mice had a modest but statistically significant decrease in lung CFUs following challenge with both A . fumigatus strains ( Fig 6C ) . Extrapulmonary dissemination was observed but did not significantly differ when comparing BALB/c and ΔdblGATA-1 mice ( Fig 6C ) . The paradoxical decreased number of CFUs in the lung in the ΔdblGATA-1 mice suggested that the mice could be succumbing to a robust inflammatory response rather than uncontrolled infection . To study this further , we looked at lung histopathology in BALB/c and ΔdblGATA-1 mice 2d following infection with the AF293 and CEA10 strains of A . fumigatus ( S3 Fig ) . In both groups , H&E stained sections revealed inflammatory cells in the airways and mainly nearby airspaces . The inflammatory cells were predominantly composed of histiocyctes admixed with sparse lymphocytes and neutrophils . Eosinophils were also noted in the lung sections from the BALB/c but not ΔdblGATA-1 mice . Gomori's methenamine silver ( GMS ) stained sections demonstrated numerous conidia and hyphae in the inflamed areas of the lungs . Interestingly , the lungs from CEA10 infected wild-type and ΔdblGATA type mice revealed relatively more hyphal growth compared to those infected with AF293 . However , despite the great difference in fungal germination , histology at two days post-infection did not demonstrate discernable differences in the inflammatory response when comparing the wild-type and eosinopenic mice infected by the two A . fumigatus strains ( S3 and S4 Figs ) . As eosinophils are thought to drive many of the pathological findings in allergic asthma , and as increased levels of IL-17 have recently been correlated with severity of asthma symptoms [35] , we next assessed whether eosinophils can also produce IL-23p19 and/or IL-17 in murine models of allergic asthma . Two established models of acute allergic asthma were chosen . In the first , mice were sensitized and challenged with A . fumigatus crude protein extracts ( Af cpe ) [36] . In the second model , sensitization was achieved with OVA admixed with aluminum hydroxide and challenge was performed with aerosolized OVA [37] ( Fig 7A ) . In each of the allergic asthma models , eosinophils produced both IL-23p19 and IL-17AF ( Fig 7B ) . In Af cpe sensitized and challenged mice , a significantly higher proportion of lung leukocytes were identified as eosinophils when compared to the OVA asthma model . Despite this difference , in both models a similar absolute number of eosinophils were recruited to the lungs ( Fig 7C ) . Greater than 90% of IL-23p19+ and IL-17AF+ cells were identified as eosinophils regardless of sensitizing allergen ( Fig 8A and 8B ) . On average , 35 . 1% of lung cells produced IL-23p19 in the A . fumigatus asthma model , compared to 22 . 5% in the OVA model ( Fig 8A and 8C ) . Sensitization with A . fumigatus cpe elicited also elicited a higher percentage of IL-17AF+ cells compared with OVA sensitization ( Fig 8B and 8D ) . Lung macrophages , inflammatory monocytes and neutrophils also produced IL-17AF , although at levels below those seen in eosinophils ( S5 Fig ) . However , among the myeloid populations studied , only eosinophils were significant producers of IL-23p19 ( S6 Fig ) .
The data presented here establish a new paradigm in acute and allergic aspergillosis whereby eosinophils act not only as effector cells but also as immunomodulatory cells driving the IL-23/IL-17 axis and contributing to inflammatory cell recruitment . Thus , following acute challenge with A . fumigatus conidia , eosinophils associated with and killed A . fumigatus conidia and ΔdblGATA-1 mice deficient in eosinophils were hypersusceptible to invasive aspergillosis . In acute and allergic models of aspergillosis , pulmonary eosinophils were found to be prominent sources of IL-23p19 and IL-17 . Finally , following acute A . fumigatus challenge , ΔdblGATA-1 mice had fewer total leukocytes , inflammatory monocytes and lung macrophages compared to wild-type mice . Although over 95% of the eosinophils were found to be positive for IL-23p19 and IL-17A ( Figs 1 and 2 ) , only 18% and 8% of eosinophils were found to associate with conidia one and three days post-infection , respectively ( Fig 4 ) . This disparity suggests that direct association with A . fumigatus is not necessary for the production of IL-23p19 or IL-17A . The milieu in which the eosinophils are stimulated appears to be important as co-incubation of bone-marrow derived eosinophils with A . fumigatus conidia or zymosan failed to elicit IL-23 or IL-17 production as measured by ELISA and ICS . These findings underscore the plasticity of eosinophils; as with other immune cells , subsets appear to exist . For example , in the small intestine , eosinophils constitutively secrete IL-1 receptor antagonist which downregulates TH17 cells [38] whereas in response to parasite challenge eosinophils promote TH2 immunity [39] . Future work should focus on determining the extrinsic and intrinsic molecular pathways that stimulate eosinophils to produce IL-23 and IL-17 . IL-23p19 and IL-17 production was observed in pulmonary eosinophils recruited to the lungs not only in models of acute and allergic aspergillosis but also in a model of allergic asthma induced by sensitization to OVA ( Figs 7 and 8 ) . As eosinophils have been reported to express IL-23R [40] , this suggests an autocrine amplification loop may be occurring whereby eosinophils release IL-23 which directly feeds back on itself via IL-23R to enhance IL-17 production . As a precedent , eosinophils express IL-5R and have IL-5 in their granules [40] . Other IL-23-producing cell types , including alveolar macrophages , have been found to produce IL-17 under defined conditions [20 , 41] . We postulate that eosinophils may be contributing to the IL-23/IL-17 axis under a broad spectrum of disease states . In support of this hypothesis , three other groups have reported evidence for eosinophil-derived IL-17 [42–43] . Using a reporter IL-17A-EGFP mouse line , Shimura et al . demonstrated that EGFP+ eosinophils in the peritoneum 6 hours following an IP LPS injection [43] . In humans , Molet et al . demonstrated IL-17 gene and protein expression in eosinophils obtained from the sputum , bronchoalveolar lavage fluid and blood of asthmatic subjects [42] . Lastly , Kobayashi et al . detected IL-17 in the supernatants of human eosinophils stimulated with monosodium urate crystals [44] . IL-23 is a key cytokine promoting and maintaining IL-17-producing cells , particularly TH17 cells [11 , 45] . Remarkably , whereas dendritic cells and macrophages are thought to be the primary sources of IL-23 in tissue [45 , 46]; in our aspergillosis models we found eosinophils were the major cell type in the lungs producing IL-23 . Thus , in the acute aspergillosis model and both allergic models , the cells that stained the brightest for IL-23p19 were nearly all eosinophils . Their identity as eosinophils was confirmed by sorting and by the loss of this population in the ΔdblGATA-1 mice . Significantly , following challenge with A . fumigatus conidia , levels of the functional IL-23 heterodimer were dramatically reduced in the BALF of mice that lacked eosinophils compared to wild-type controls . Two IL-23p19 antagonists , tildrakizumab and guselkumab , are undergoing clinical trials for the treatment of autoimmune and allergic disease [46] . For the subset of these diseases in which eosinophilia is seen , our data suggest that targeting eosinophils could have therapeutic efficacy while avoiding potential toxicities such as impaired host defenses to infections that could occur following global inhibition of the IL-23/IL-17 axis [46] . In mouse models of asthma , the absence of eosinophils correlates with decreased airway hyperresponsiveness ( AHR ) and decreased mucus production [47]; these findings are also found in mice lacking IL-17 signaling [48 , 49] . Another example of functional overlap is the finding that IL-17 signaling and eosinophil co-culture with epithelial cells have each been shown to up-regulate Mucin 5AC , a mucin protein produced by bronchial airway goblet cells [48 , 50 , 51] . The convergence between IL-17 and eosinophils found in our models of allergic asthma point to a novel mechanism by which eosinophils contribute to the pathogenesis of asthma in a manner that is independent of TH2-driven inflammation . If this phenomenon holds up in humans , as is supported by the data of Molet et al . [42] , it could be of use to better stratifying asthma type , as heterogeneity in this disease has prevented the success of some targeted asthma therapies [48 , 52] . Clinically , asthma features episodic shortness of breath accompanied by wheezing due to AHR . However , it has been increasingly recognized that a heterogeneous group of immunological processes drives these hallmark symptoms [48 , 53] . Individualizing therapy based on the immunological drivers has the potential to improve therapeutic outcome , which is relevant considering many patients fail to respond to standard treatment [53 , 54] . Stratification based on the predominant cellular infiltrate found in induced sputum has delineated four distinct types of asthma: eosinophilic , neutrophilic , mixed granulocytic and paucigranulocytic [55] . It will be interesting to assess the presence of IL-23+ IL-17+ eosinophils in each of these types , as even in neutrophilic and paucigranulocytic asthma a small number of eosinophils is present [55] . While eosinophils act as drivers of inflammation and tissue remodeling in asthma [39 , 52 , 56] , they play a protective role in models of acute pulmonary aspergillosis . Eosinopenic mice challenged with a relatively virulent strain of A . fumigatus had increased mortality compared to wild-type mice ( Fig 6 ) despite similar CFUs and no striking differences in lung pathology ( Figs 6 , S3 and S4 ) . In contrast , Lilly et al . found Aspergillus-infected mice lacking eosinophils had higher fungal burdens , as measured by A . fumigatus 18S rRNA and fungal germination [10] . How eosinophils protect hosts still needs to be more fully defined and may be multifactorial . Eosinophils directly kill A . fumigatus in vivo ( Fig 4 ) and by contact-dependent mechanisms in vitro [10] . Moreover , lysates containing eosinophil granule proteins inhibit fungal germination [10] . However , we also found eosinophils act indirectly by promoting the recruitment of inflammatory monocytes and macrophages to the lungs after infection ( Fig 5 ) . Alveolar macrophages and inflammatory monocytes kill A . fumigatus conidia [29 , 57] . In addition , mice depleted of inflammatory monocytes are vulnerable to A . fumigatus infection [29] . Interestingly , ΔdblGATA-1 mice did not have a defect in neutrophil recruitment to the lung following A . fumigatus challenge ( Fig 5 and [10] ) . Moreover , in a neutropenia mouse model of aspergillosis that features repeated aspiration of conidia from an unusual strain of A . fumigatus containing high cell wall chitin , decreases in fungal burden and mortality were observed in ΔdblGATA-1 mice [9] . In the context of the damage-response framework of microbial pathogenesis [58] , we posit that eosinophils and eosinophilic production of IL-23 and IL-17 are beneficial in invasive aspergillosis but detrimental in allergic disease . Future studies will test this hypothesis and examine the spectrum of diseases under which eosinophils produce IL-23 and IL-17 . These data could help inform clinical studies using monoclonal antibodies and other immunomodulators that target eosinophils and the IL-23/IL-17 axis .
Six- to eight-week old C57Bl/6 , BALB/c , ΔdblGATA-1 mice in the BALB/c background , and IL-17AGFP/GFP mice in the C57Bl/6 background were obtained from Jackson Laboratories and bred in pathogen-free conditions at the University of Massachusetts Medical School ( UMMS ) . The mouse studies were performed in accordance with protocol #1802–15 approved by the UMMS Institutional Animal Care and Use Committee . In addition , all mouse studies were carried out in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . A . fumigatus from the AF293 strain was obtained from the Fungal Genetics Stock Center ( Manhattan , KS ) . AF293 was grown from frozen stocks on Sabouraud-dextrose agar ( Remel ) slants [59 , 60] . The CEA10 strain was a generous gift from Robert Cramer ( Geisel School of Medicine , Hanover , NH ) and grown as described [61] . Conidia were harvested by vortexing slants with PBS containing 0 . 01% Tween-20 ( Thermo-Fisher ) and filtering the suspension twice through a 30 μm nylon mesh folded over a conical tube . Suspensions were then washed three times with 0 . 01% Tween-PBS , and re-suspended in the same solution at a concentration of 6 . 67x108 conidia/mL . Mice were infected via the oro-tracheal ( OT ) route with 75 μL of conidial suspension , so that each mouse received 5x107 conidia . OT infection was facilitated by anesthetizing mice with isoflurane ( Piramal Healthcare ) . Organ CFUs were enumerated as in our previous studies [62 , 63] except the resected organs were homogenized and diluted in PBS containing 0 . 01% Tween-20 . The lower limit of detection was 10 CFU/organ . Bone marrow cells were isolated from murine femurs and differentiated as described previously [64] . Briefly , BM cells were subjected to a gradient with Histopaque 1083 ( Sigma ) . Interphase cells were washed twice to remove any leftover Histopaque and cultured in Isocove’s modified Dulbecco’s medium ( IMDM; Life Technologies ) supplemented with 10% FBS ( Tissue Culture Biologicals ) , 100 U/mL penicillin and 100 U/mL streptomycin , 1X Glutamax ( Life Technologies ) and 2 μL of β-mercaptoethanol ( Life Technologies ) . The first four days of culture , cells were stimulated with 100 ng/mL of murine stem cell factor ( mSCF ) ( PeproTech ) and 100 ng/mL of murine FMS-like tyrosine kinase 3 ligand ( mFLT3L ) ( PeproTech ) . The following ten days , cells were stimulated with 10 ng/mL IL-5 ( R&D ) . Cells were fed every other day . Differentiation was confirmed by flow cytometry with rat mAb against murine Siglec-F-BV421 ( BD Biosciences ) . BM-Eos were stimulated with live or heat-killed A . fumigatus AF293 conidia , zymosan ( 10–100 μg/mL; Sigma ) , lipopolysaccharide ( LPS , 100 ng/mL; Sigma ) and several cytokines at different concentrations and combinations for lengths of time ranging from two hours to two days . Cytokines used to stimulate BM-Eos included IL-5 ( 10 ng/mL ) , IL-1β ( 1–10 ng/mL; R&D ) , GM-CSF ( 10 ng/mL; PeproTech ) , IL-23 ( 10 ng/mL; eBiosciences ) , IL-17E ( 0 . 1–10 ng/mL; R&D ) , IL-17AA ( 0 . 1–10 ng/mL; R&D ) , PGE2 ( 10−5–10−3 M; Cayman Chemicals ) , transforming growth factor-β ( TGF-β ) ( 1–10 ng/mL; eBiosciences ) , and IL-6 ( 10–100 ng/mL; eBiosciences ) . When co-incubated with live A . fumigatus conidia , cultures were supplemented with 0 . 5 μg/mL of voriconazole ( Sigma ) to prevent fungal overgrowth . Following stimulation , supernatants were collected for ELISA . In some experiments , cells were co-incubated with 1 μM monensin in the last 5 hours of stimulation , and intracellular cytokine staining was performed as described below . FLARE conidia were created as described [26 , 65] . Briefly , 5x108 dsRed conidia ( AF293 ) per mL were incubated with 0 . 5 mg/mL 6- ( ( 6- ( ( Biotinoyl ) Amino ) Hexanoyl ) amino ) Hexanoic Acid , Sulfosuccinimidyl Ester , Sodium Salt ( Biotin-XX , SSE ) ( ThermoFisher ) in 50 mM carbonate buffer ( Sigma ) , pH 8 . 3 at 4°C for 2 h in a tube rotator . Excess Biotin-XX , SSE was washed off with 0 . 1 M Tris-HCl pH 8 . 0 ( Sigma ) ; then conidia were incubated with 0 . 02 mg/mL streptavidin conjugated to Alexa Fluor 633 ( Life Technologies ) away from light at room temperature for 30 minutes . Labeling was confirmed by flow cytometry prior to infecting animals . C57Bl/6 mice were sensitized two times by IP injection of 20 μg of OVA ( Sigma-Aldrich ) in 100 μL of Imject Alum ( Thermo Scientific ) or 200 μg of A . fumigatus crude protein extracts ( Af cpe ) ( Greer Laboratories ) . Sensitization occurred two weeks apart . On days 28 , 29 and 30 mice were challenged with aerosolized 1% OVA or 0 . 25% Af cpe in saline respectively for 20 minutes as described [36] . ICS was performed as previously described [66] . Briefly , mice were treated with 500 μg of monensin ( Sigma-Aldrich ) by IP injection , 2 or 48 h after infection with 5x107 conidia , or 48 h after the last challenge in the allergic asthma models . Six hours after monensin treatment , lung single-cell suspensions were prepared using MACS lung dissociation kit as described by the manufacturer ( Miltenyl Biotec ) . Single-cell suspensions were enriched for leukocytes using a Percoll ( GE Healthcare ) gradient [67] . Interphase cells were collected , counted with the aid of a hemocytometer , and then co-incubated with rat anti-mouse CD16/CD32 monoclonal antibody ( mAb ) 2 . 4G2 ( BD Pharmingen ) to block Fc receptors in accordance with manufacturer’s directions . Surface antigens were then stained with antibodies listed in Table 1 and with Fixable Viability Dye eFluor 780 ( eBioscience ) or Live/Dead Blue ( Life Technologies ) for 30 minutes at 4 C . After two successive wash steps , lung leukocytes were fixed in 2% paraformaldehyde ( Electron Microscopy Sciences ) PBS solution for at least 15 minutes at 4 C . Fixed cells were permeabilized using Perm/Wash Buffer ( BD Pharmingen ) according to manufacturer instructions and then stained with antibodies listed in Table 1 . For sorting IL-23p19+ IL-17A+ cells , fixed lung leukocytes were only stained for intracellular cytokines , and then sorted with BD FACSVantage DV-1 Cell Sorter ( UMass Medical School Flow Cytometry Core ) . FC data were acquired with a BD LSR II cytometer and analyzed using FlowJo X software ( Tree Star Inc . ) . Gating was established using FMO controls containing isotype control mAb conjugated with the fluorophore corresponding to the missing antibody as described [68] . Mice were infected with 5 x 107 A . fumigatus strain 293 conidia via oral-tracheal route . After two or 48 h , mice received 500 μg of IP monensin , and 6 h later the mice were euthanized and bronchoalveolar lavage fluid was collected . Cells were washed with PBS , counted , fixed with 2% paraformaldehyde buffered in PBS and washed an additional 3 times . The lavage cells were then cytospun onto slides ( Shandon Cytoslide ) , permeabilized for 10 minutes in 0 . 1% Triton-X 100 in PBS , washed three times with PBS and blocked with 10% donkey serum in PBS for 1 h at room temperature . Cells were sequentially immunostained in 5% donkey serum with goat polyclonal antibody ( 4°C , 3 hours ) , Alexa fluor 594-labeled donkey anti-goat fab2 ( room temperature , 1 hour ) , rabbit polyclonal antibody ( 4 hours , 4°C ) , and Alexa fluor 647-labeled donkey anti-rabbit fab2 fragment ( 45 minutes , room temperature ) . The specific antibodies used , including the control antibodies , are listed in Table 2 . Cell nuclei were stained with 4’6-diamino-2-phenylindole ( DAPI , 1ug/ml; Sigma ) and the slides were mounted with ProLong Gold Antifade Mountant ( Life Technologies ) . Images were acquired with a 63x oil immersion objective at maximum intensity projections in sequential scan mode with a laser confocal microscope ( Leica SP8 ) . After sorting , cells were adhered to poly-L-lysine-coated slides ( Sigma-Aldrich ) by cytospinning 250 μL of cell suspension at 800 rpm for three minutes ( Shandon Cytospin 2 ) . Cells were dried on the slide , and then dipped in hematoxylin solution ( Thermo-Fisher ) for 30 seconds . Slides were washed in water then dipped in eosin ( Thermo-Fisher ) for one minute . Increasing concentrations of ethanol ( 95%-100% ) were used to dehydrate slides for a total of two minutes . Finally , slides were dipped three times in Clear-Rite ( Thermo Scientific ) for one minute each time . Slides were mounted using Permount ( Fisher Scientific ) . BALF from BALB/c and ΔdblGATA-1 mice infected as described above was collected after euthanasia at 60 h post-infection . Briefly , an 18 or 20 gauge plastic catheter ( Temuro ) was inserted into the trachea and lungs were flushed three times with 1 mL of PBS supplemented with cOmplete protease inhibitor cocktail ( Sigma-Aldrich ) . BALF samples were flash frozen in dry ice and subsequently stored in -80°C . IL-23 heterodimer ( IL-23p19/IL-12p40 ) was measured using R&D Quantikine ELISA kit . Statistical tests were performed using GraphPad Prism 6 . The Student’s t-test was used to compare the means of two groups with the Bonferroni correction applied for multiple comparisons . For data sets where the mean of more than two groups was compared , 2-way ANOVA was performed using Tukey’s multiple comparison correction . In comparing Kaplan-Meyer survival curves , the Mantel-Cox test was used .
|
The opportunistic fungus , Aspergillus fumigatus , causes a spectrum of diseases ranging from invasive aspergillosis in the severely immunosuppressed to allergic asthma in atopic individuals . Here we explored the contribution of eosinophils , a type of white blood cell , to host defenses and pathogenesis in murine models of invasive pulmonary aspergillosis and asthma . We found eosinophils co-produce the cytokines IL-23 and IL-17 in both aspergillosis models as well as a model of OVA-induced asthma . Eosinophils killed the conidia ( spores ) of A . fumigatus in vivo and mice that lacked eosinophils were more susceptible to invasive aspergillosis . These observations suggest eosinophils play a more prominent role in defenses against invasive pulmonary aspergillosis than heretofore appreciated and identify eosinophil-derived IL-23 and IL-17 as potential therapeutic targets in allergic asthma .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"blood",
"cells",
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"aspergillus",
"fumigatus",
"immune",
"cells",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"cytokines",
"aspergillus",
"pathogens",
"immunology",
"microbiology",
"pulmonology",
"animal",
"models",
"developmental",
"biology",
"fungi",
"model",
"organisms",
"signs",
"and",
"symptoms",
"experimental",
"organism",
"systems",
"molecular",
"development",
"fungal",
"diseases",
"aspergillosis",
"fungal",
"pathogens",
"asthma",
"research",
"and",
"analysis",
"methods",
"specimen",
"preparation",
"and",
"treatment",
"infectious",
"diseases",
"mycology",
"white",
"blood",
"cells",
"inflammation",
"staining",
"animal",
"cells",
"medical",
"microbiology",
"microbial",
"pathogens",
"mouse",
"models",
"molds",
"(fungi)",
"immune",
"response",
"immune",
"system",
"cell",
"staining",
"eosinophils",
"diagnostic",
"medicine",
"cell",
"biology",
"physiology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"organisms"
] |
2017
|
Central Role of IL-23 and IL-17 Producing Eosinophils as Immunomodulatory Effector Cells in Acute Pulmonary Aspergillosis and Allergic Asthma
|
Visceral leishmaniasis ( VL ) is an important neglected disease caused by a protozoan parasite , and represents a serious public health problem in many parts of the world . It is zoonotic in Europe and Latin America , where infected dogs constitute the main domestic reservoir for the parasite and play a key role in VL transmission to humans . In Brazil this disease is caused by the protozoan Leishmania infantum chagasi , and is transmitted by the sand fly Lutzomyia longipalpis . Despite programs aimed at eliminating infection sources , the disease continues to spread throughout the Country . VL in São Paulo State , Brazil , first appeared in the northwestern region , spreading in a southeasterly direction over time . We integrate data on the VL vector , infected dogs and infected human dispersion from 1999 to 2013 through an innovative spatial temporal Bayesian model in conjunction with geographic information system . This model is used to infer the drivers of the invasion process and predict the future progression of VL through the State . We found that vector dispersion was influenced by vector presence in nearby municipalities at the previous time step , proximity to the Bolívia-Brazil gas pipeline , and high temperatures ( i . e . , annual average between 20 and 23°C ) . Key factors affecting infected dog dispersion included proximity to the Marechal Rondon Highway , high temperatures , and presence of the competent vector within the same municipality . Finally , vector presence , presence of infected dogs , and rainfall ( approx . 270 to 540mm/year ) drove the dispersion of human VL cases . Surprisingly , economic factors exhibited no noticeable influence on disease dispersion . Based on these drivers and stochastic simulations , we identified which municipalities are most likely to be invaded by vectors and infected hosts in the future . Prioritizing prevention and control strategies within the identified municipalities may help halt the spread of VL while reducing monitoring costs . Our results contribute important knowledge to public and animal health policy planning , and suggest that prevention and control strategies should focus on vector control and on blocking contact between vectors and hosts in the priority areas identified to be at risk .
Visceral leishmaniasis ( VL ) , also known as kala-azar , is characterized by irregular bouts of fever , weight loss , enlargement of the spleen and liver , and anemia [1] . It is an important disease that occurs around the world [2] , and Brazil is one of six countries that concentrates 90% of new human VL cases [1] . In Europe and Latin America , the disease is zoonotic , and in Brazil it is caused by the parasite Leishmania infantum chagasi ( L . i . chagasi ) , which is transmitted to humans and other animals ( such as rodents and wild and domestic dogs ) mainly through the bites of a sand fly ( Lutzomyia longipalpis ) vector [3] . Despite being historically known as a rural endemic disease , VL has reached endemic and epidemic proportions in many large Brazilian cities since the 1980s [3 , 4 , 5] , representing a serious public health problem [2] . Where the disease is zoonotic , infected dogs constitute the main domestic parasite reservoir and play a key role in VL transmission to humans [6 , 7] . Hence , one of the main VL prevention and control strategies adopted by the Brazilian government has been to test and cull infected dogs [8] . Despite these efforts , ongoing program initiatives to combat infection sources have failed to reduce disease incidence in humans [9] . On the contrary , the number of new notified cases has kept rising and the disease is spreading across Brazil [10 , 11]] . The Brazilian State of São Paulo has experienced rapid spread of human and dog VL over the last two decades , mainly associated with the expansion of the vector’s geographic range . In 1998 , the sand fly that transmits VL was found in only two municipalities , but since then it has been recorded in 164 municipalities by 2014 , principally in the western part of the State [12] . In 1998 and 1999 the first infected dog [13] and human cases were identified , respectively , in municipalities where the vector had already been confirmed [14] . These changes coincided with the construction of the Bolívia-Brazil gas pipeline , which recruited large numbers of migrant workers from various VL endemic regions ( e . g . , Mato Grosso do Sul ) . The pipeline was built across São Paulo State between 1997 and 1999 , and within ten years , 53 São Paulo municipalities confirmed 954 human cases of VL and reported 81 associated deaths between 2010 and 2015 [10 , 15] . The expansion of VL in São Paulo occurred along a major axis extending from the northwest to the southeast towards the Bauru region , following both the Bolivia-Brazil gas pipeline and the Marechal Rondon Highway [16] . Migrant workers arrived during the construction of the gas pipeline by way of the Marechal Rondon Highway , which crosses from the northwest corner of the state to the state capital in the southeast . Other factors may also have contributed to VL expansion , such as socioeconomic conditions , climatic factors and ecological imbalances that influence vectors responsible for transmitting the L . i . chagasi parasite . For instance , Sundar et al . [17] found an association between poverty and VL incidence in humans , and suggested that environmental factors created by poor housing conditions encourages sand fly breeding . Nevertheless , the ultimate factors that contribute to the spread of VL remain uncertain . Infectious disease outbreaks have both spatial and temporal dimensions [18] . Spatial heterogeneities may arise due to differences in the distribution of vectors and other risk factors , different patterns of intra-urban vector-host contact , and variations in population susceptibility [5] . Few studies have assessed the spatio-temporal distribution , including underlying risk factors , of VL [19–22] , particularly in Brazil [23–25] . Furthermore , to the best of our knowledge , this is the first study to predict the spread of VL through time and space in São Paulo State , where the disease now seems to have reached the capital . Unlike other studies [23 , 24] , we employ an innovative spatio-temporal Bayesian model and GIS ( Geographical Information Systems ) to infer factors driving the invasion of VL by integrating data on hosts and vector locations between 1999 and 2013 with disease natural history . This Bayesian model is then used to predict the epidemic’s future geographical progression . Prediction maps identify strategic intervention areas for controlling vector and infected host dispersion , and provide valuable information to public health authorities to help spatially optimize VL prevention and control measures .
São Paulo State is comprised of 645 municipalities ( Fig 1 ) , with geographic areas ranging from 118 , 951 and 1 , 977 , 951 km2 . Based on digitized municipal boundary maps produced by the Brazilian Institute of Geography and Statistics [26] , we calculated geographic centroids for each Municipality using QGIS ( version 2 . 8 ) , which are then used in subsequent analyses . For each Municipality , we obtained data on the presence and absence of ( 1 ) VL sand fly vectors , ( 2 ) autochthonous VL human cases , and ( 3 ) VL seropositive dogs , from 1999 to 2013 . Data on the distribution of the vector Lu . longipalpis were obtained from the São Paulo State Secretary of Health , as described by Casanova et al . [12] . The notification of human VL cases is compulsory throughout Brazil , and is monitored through ( 1 ) spontaneous reporting to the Brazilian Health Units , ( 2 ) active search for cases in known transmission areas , ( 3 ) home visits by health professionals , and ( 4 ) patient referrals through the primary health care network [27] . We assume these records to be the best available for the study area . They are freely available through the Brazilian Ministry of Health website [10 , 11 , 15 , 28] . Data on infected dog cases were obtained from canine surveys carried out by municipal governments , the Adolfo Lutz Institute , and the Secretary of Health of São Paulo State . Municipalities are also expected to notify the first-confirmed occurrence of L . i . chagasi in dogs . Data from municipalities arise from dogs suspected to be infected that are subsequently evaluated using a parasitological test . To investigate factors driving the spread of sand fly vectors , VL infected dogs and infected human dispersion across São Paulo State , we examined the following covariates at the municipal level: invasion pressure , climatic features , economic factors , and the distance between each municipality’s centroid and ( 1 ) the Marechal Rondon Highway and ( 2 ) the Bolivia-Brazil gas pipeline . Bayesian models for disease mapping have primarily utilized a generalized linear model framework , with regression parameters modeled as random effects with spatial ( or space-time ) covariance matrix [34] . Specific studies that applied these models to leishmaniasis in Brazil include Karagiannis-Voules et al . [23] and Assunção et al . [24] . The Bayesian model we developed to predict spatio-temporal spread of VL in São Paulo State differs substantially from these earlier approaches in that our model integrates data on vectors , infected dogs and humans , and accounts for spatial and temporal correlation through a spatially weighted infection pressure index , as described below . Our observations consist of yit , which is a binary outcome indicating whether municipality i was invaded in year t , where i = 1 , … , 645 , and t = 1 , … , 15 . The time subscript t corresponds to annual time steps from 1999 to 2013 . In all of our models , we assumed that yit~Bernoulli ( θit ) where the probability of invasion is given by θit If a municipality has already been invaded by a competent vector ( or by an infected dog or human ) , we assume it will keep its invasion status throughout the study period . This assumption can be stated as: p ( yit=1|yi ( t−1 ) =1 ) =θit=1 If the municipality has not been previously invaded ( i . e . , yi ( t-1 ) = 0 ) , we assume that: p ( yit=1|yi ( t−1 ) =0 ) =θit=exp ( xitTβ ) 1+exp ( xitTβ ) where xitT and β are vectors of covariates and regression parameters , respectively . We base our investigations of municipality “invasion” on the following premises , related to the competent vector and infected hosts: Finally , we assume no false positives or false negatives . The models formulated above can be thought of as generalizations of the geometric distribution ( number of trials before the first success/invasion ) , where the individual success/invasion probabilities are not constant and spatial and temporal dependence are taken into account by evaluating the status of other municipalities in the previous time step , as captured by the invasion pressure covariate for vectors , infected dogs and infected humans ( Pitv , Pitd , Pith respectively ) . Our analysis is conditional on the status of each municipality in year 1 . Thus , the likelihood of this model is given by: p ( y . 2 , … , y . T|X , β , y . 1 ) ∝∏i=1I∏t=2Tp ( yit=1|y . ( t−1 ) , xit , β ) We fit this model in a Bayesian framework using JAGS , which enables us to coherently represent uncertainty when creating predictions of invasion probability from 2014 to 2020 . Our Bayesian model is likely to be very useful for researchers interested in understanding disease spread , and for this reason , we include a detailed tutorial-style appendix that describes how to implement the model ( S1 Appendix ) . The model described above was also used to predict the “invasion” of VL-free municipalities from 2014 to 2020 by using the posterior distribution of the parameters and forward simulation of the invasion process . For each sample of the posterior distribution , we sequentially created a forward simulation for the vector , for the infected dogs , and finally for infected humans , using 2013 as the starting point . We then summarized these synthetic invasion scenarios by calculating the probability that each municipality has been invaded . Our annual predictions of these invasion probabilities were mapped ( using QGIS version 2 . 8 ) , and maps were compiled into a time-loop video ( available online ) that incorporates both actual vector and infected host presence/absence data from 1999–2013 as well as predicted probability of invasion from 2014–2020 . For model validation , we obtained 2015 data on municipalities invaded by the vector and VL infected humans . Model validation was accomplished by comparing the 2015 predicted invasion probabilities with actual 2015 observations . Although only five and 15 new municipalities were invaded by infected humans or vectors in 2015 , respectively , these data were still useful in evaluating the model’s predictive power .
A summary of our invasion risk factor analysis for vectors , dogs and humans is given in Table 1 . Statistically significant covariates for vector dispersion included the invasion pressure index ( which takes into account vector presence in nearby municipalities during the previous time step ) , proximity to the gas pipeline , and temperature . Covariates that significantly affected the dispersion of infected dogs were proximity to Highway , temperature , and presence of the competent vector within the same Municipality . Finally , the covariates that exhibited significant influence on the dispersion of infected humans were vector presence and the interaction term between vectors and infected dogs . The effect of rainfall was also marginally significant for human cases of VL . Both vector and infected dog dispersions were strongly influenced by temperature . On the other hand , none of the dispersion patterns were affected by altitude variations or low economic productivity ( GDP ) . Interestingly , results regarding proximity to the gas pipeline and the highway were inconsistent , suggesting potentially different dispersion mechanisms for vectors versus dogs . Maps of vector , infected dog , and infected human presence/absence for 2013 and prediction maps for 2016 , 2018 , and 2020 displaying “invasion” probabilities are shown in Fig 2 . The effect of distance to the Highway on infected hosts is particularly evident in Fig 2 . Vectors are spreading toward the north and east where municipalities intersect the gas pipeline , and therefore , these regions are associated with a higher probability of vector invasion ( Fig 2 and Table 1 ) . As a result , both infected dog and human invasion probabilities in municipalities crossed by the gas pipeline are also slightly higher , due to the influence of the vector in their dispersion ( Table 1 ) . The majority of municipalities currently affected or predicted to be affected by vectors or hosts lie in the western part of São Paulo State . The top 20 municipalities with the highest invasion probabilities for vectors , infected dogs and infected humans are displayed in Fig 3 and listed in S1 Table . As expected for infected dogs , municipalities with higher invasion probabilities in 2020 were typically those already invaded by the vector ( Fig 2 ) . Three municipalities ( enumerated as 33 , 35 and 37 in Fig 3 ) have both vectors and infected dogs , and no infected humans ( at least until 2013 ) ( see in Fig 2 ) . For infected humans , municipalities with the highest invasion probabilities in 2020 were those already invaded by both vectors and infected dogs . Three municipalities ( enumerated as 5 , 14 and 34 in Fig 3 ) have both infected dogs and humans , as well as the vector ( see in Fig 2 ) , which “arrived” before 2013 . As expected , our models tended to assign higher invasion probabilities to “invaded” municipalities than to “non-invaded” municipalities , indicated by the taller boxes for the “invaded” municipalities in Fig 4 . Furthermore , the difference in predicted probabilities for the invaded and non-invaded municipalities was greater for VL human cases than for vector presence ( Fig 4 ) , suggesting a higher predictive skill of our models for VL human cases rather than vector presence . This is likely due to the fact that vector and infected dog presence in 2013 is highly predictive of VL human case invasion in 2015 . Indeed , two out of five municipalities invaded with VL human cases in 2015 were already invaded by infected dogs by 2013 , and all five municipalities had already been invaded by the vector in 2013 .
One of the most important risk factors for VL around the world is the migration of people from endemic to non-endemic regions [37] . During the construction of the Bolivia-Brazil gas pipeline , which began in 1998 , thousands of workers moved from the city of Corumbá ( an endemic area ) to other cities and regions of Mato Grosso do Sul and neighboring states ( all non-endemic areas ) , explaining the spatial pattern of disease establishment in this region [38 , 39 , 12] . Cardin et al . [16] were the first to observe a similar pattern in São Paulo State , finding an association between leishmaniasis and proximity to the Marechal Rondon Highway . They reasoned that the road served as an important conduit for arrival of migrants into São Paulo , connecting endemic and non-endemic regions [16] . The Marechal Rondon Highway provided such a link from western municipalities in the State to the capital city . We found a significant influence of the gas pipeline on VL distribution , specifically on vector dispersion , and an effect of the Highway on the dispersion of infected dogs ( see Table 1; note that the two coefficients are negative , suggesting that the smaller the distance , the greater the probability of invasion ) . While VL presence may have been initially influenced by the gas pipeline construction , it may now be spreading along major transport routes such as the Highway , as observed in our predictions ( Fig 2 ) . In their study of a leishmaniasis outbreak in Spain , Ruiz et al . suggested that construction of a dense road network created ideal conditions for the development of suitable habitats for both VL vectors and vector reservoirs [21] . In the same way , the Marechal Rondon Highway's presence in São Paulo altered environmental conditions by removing the native vegetation , and likewise , the construction of the gas pipeline may have prompted an increase of VL incidence by altering vector habitat . We found no significant effect of altitude on vector or infected host dispersion in São Paulo State . Vector and infected hosts occurred in municipalities with elevation between 274m and 804m , but preferentially between 274m and 539m , representing 85 . 3% and 40% of the municipalities altitudes at the State , respectively ( Fig 2 and S3 Fig ) . The vector itself was present in just a few municipalities between 804 and 1040m . In Belo Horizonte , Brazil , Etelvina et al . [40] also observed no influence of altitude in human VL cases; however , other studies conducted in the same area suggested a concentration of infected dogs and human VL cases between 780 and 880 m [41] , or between 751m and 850m of altitude [42] , and found that the majority of sand fly vectors occur at altitudes below 851m [42] . Factors such as temperature , relative humidity , and rainfall can influence sand fly population density [43] . In the present study we observed that higher temperatures ( annual average between 19 and 23°C ) , occurring in the western part of the State ( S4 Fig ) , favor the spread of vectors ( Table 1 ) . Furthermore , in our predictions ( Fig 2 ) , neither vectors nor infected hosts reached areas cooler than 17°C , typical of municipalities on the eastern side of the State ( S4 Fig ) . Other studies revealed that temperature strongly influences the distributions of competent VL vectors and infected dogs in Brazilian endemic areas [44 , 45] , as well as in parts of Europe [46 , 47] . In studies conducted over smaller spatial scales , where temperatures vary little throughout the year , temperature was found to have no effect on vector dispersion [48 , 49] . Temperature regulates several sand fly biological parameters [50] . In particular , low temperatures reduce the sand fly metabolism ( thus increasing its longevity ) [51 , 52] and biting rate , and also increase extrinsic incubation period , thereby negatively affecting the sandflies’ basic reproduction rate [50] . Hence , climate change , especially an increase in temperature , may accelerate the rate of geographical expansion of VL . A map of average rainfall ( S5 Fig ) shows that dry areas ( annual average as 96-119mm/year ) predominate in the western part of São Paulo State , representing 69 . 3% ( 447/645 ) of all municipalities . Between 1998 and 2013 , vectors and infected dogs were concentrated mainly in these western municipalities , and rainfall had no significant effect on their distributions ( Table 1 ) . On the other hand , rainfall affects the infected human dispersion ( see Table 1; the negative coefficient suggests that lower rainfall increases the probability of invasion ) , and 89% of the municipalities with human VL cases experience an average rainfall of 96-109mm/year , and represent 42 . 8% ( 276/645 ) of all São Paulo municipalities . Studies suggest that dry seasons are associated with increased numbers of VL canine cases [53] as well as greater vector densities in Mato Grosso do Sul-Brazil [53 , 54] and Costa Rica [55] . In Spain , Pérez-Cutillas et al . [56] observed that areas with high human VL prevalence had significantly less rainfall during the sand fly season and more rainfall in wintertime . In addition , these authors suggested that greater rainfall during the coldest months of the year may increase survival of eggs and diapausing sand fly instars , while greater precipitation during the summer months may hamper adult sand fly flight activity . In Brazil , other environmental factors have been positively correlated with anti-Leishmania antibodies in humans , including houses with mud floors and/or mud walls , lack of garbage collection , vegetation type , and presence of dogs , birds , or donkeys/horses in the neighborhood [57] , as well as inadequate sanitation infrastructure [58] . These factors likely reflect an overlap of poorer housing with other unmeasured effects of poverty and/or an environment conducive to sand fly breeding [57] . High indices of vector diseases are often correlated with the world’s poorest regions , and low socioeconomic status of dog owners is associated with L . i . chagasi canine infection [35 , 59–63] . Lima et al . observed that L . i . chagasi infections in dogs and humans were intimately connected to environmental conditions associated with low incomes [57] . Karagiannis-Voules et al . found a direct association between socioeconomic factors and the incidence of human VL cases in Brazil [64] . In the present study , however , vector presence and infected host dispersions presented no significant relationship with income as measured by GDP . GDP varied only slightly among São Paulo municipalities where vectors and infected hosts were found ( Fig 2 ) , which may explain why we found no association . More stark differences in economic productivity ( GDP ) are evident in the eastern region of the State , but this region has displayed very low VL invasion up till now ( Fig 2 ) . Finally , better poverty metrics might present a stronger correlation with dispersion probability of vectors and infected hosts . We originally hypothesized that the invasion pressure index would play a key role in dispersion of vectors , infected dogs and infected humans , but our results only showed an impact on vector dispersion , suggesting that spatial contagion is a key process only for vectors . The spatial contagion ( as captured by our invasion pressure covariate ) was not significant , given the presence of the vector , likely due to the relatively high mobility of dogs and humans , as compared to the vector . These results highlight the importance of simultaneously accounting for vectors and multiple hosts when assessing the spatial and temporal distribution of human VL cases . Presence of the vector itself is a poor predictor of disease transmission , because infected reservoirs or/and vectors must be present to establish parasite circulation in transmission foci [65] . Nonetheless , spatially targeted interventions aimed at eliminating the vector could be highly effective in halting vector dispersion , whereas similar spatial strategies targeting infected dogs or humans are less likely to succeed . Prioritizing insecticide spraying of residual action and other strategies that aim to eliminate or avoid VL vectors in municipalities with a high invasion probabilities represent an integral and necessary component of preventative measures for this disease . In Brazil some studies found that the occurrence of VL in humans is associated with the presence of infected dogs in the same area [25 , 26 , 49 , 50 , 66 , 67] . Werneck et al . observed that increasing prevalence of canine infection effectively forecast high incidence of human VL , as did high prevalence of canine infection before and during an epidemic [60] , and Lima et al . report a high likelihood of geographic intersection between seropositive humans and dogs [57] . In the present study , infected dogs also affect the dispersion of infected humans , but only when the competent vector is also present in the same environment . This particular dynamic occurs where the disease presents a zoonotic pattern , with dogs acting as the main reservoir . Considering this zoonotic scenario , and in an effort to reduce or even eliminate human cases of VL , the Brazilian Leishmaniasis Control Program treats human cases and invests in the reduction of vector densities ( spraying insecticide ) and canine control ( culling seropositive dogs ) [27] . Nevertheless , these strategies , practiced since 1950 , have met with little success in reducing VL incidence in humans . Instead , the number of human VL cases continues to rise , and the disease is spreading geographically becoming a serious public health problem [9] . Because the presence of the vector and infected dogs together affected the dispersion of human VL cases , we argue that while control strategies should continue to focus on vector-dog contact , inclusion of other measures that have demonstrated success in controlling the disease should also be promoted , such as dog vaccines [68 , 69] insecticide-impregnated collars [13 , 70–73] for a large percentage of dog population [74] , application of insecticide of residual action in the environment , topically applied insecticide , and testing dogs before their introduction into non-endemic areas [75] . Other models used to predict zoonotic VL [19 , 23 , 65] do not integrate data on the vector and the two hosts ( i . e . , dogs and humans ) . Furthermore , unlike these other studies , we map annual probabilistic predictions for the vector , infected dogs and infected humans using GIS . Along with these innovations , we acknowledge important limitations regarding the data used in our statistical analysis . First , leishmaniasis cases are often under-reported in Brazil [76 , 16] . For this reason , we restricted our analysis to presence/absence data and assumed that once “invaded” , a municipality remains “invaded” , even if no additional cases are detected . Second , the aggregate nature of our data ( i . e . , the smallest unit of analysis was the municipality ) may have masked associations with other factors ( e . g . , the influence of socio-economic conditions on the presence of the vector and infected hosts ) . We used a Bayesian model to better define the risk factors for the spread of VL in São Paulo State , integrating data on its vector and hosts , and to predict the distribution and dispersion of the disease over time and space . As we already cited above , zoonotic VL has been spreading in other regions of Brazil outside of São Paulo [77 , 24 , 38] and in other countries as well [21] . Extension of our model to other scenarios can offer further insights into the factors influencing the spread of leishmaniasis and other diseases elsewhere .
Landscape features , such as the gas pipeline and Marechal Rondon Highway and annual temperature , rather than economic factors , were the most important risk factors in predicting VL dispersion in São Paulo State , Brazil . Since the dispersion of infected humans is affected by the spatial distribution of vectors and infected dogs , strategies to block the spread of the disease in humans and dogs need to address all three components of the VL dynamic cycle . Prevention and control measures should not only focus on vector control ( e . g . , use of residual insecticide ) , but also employ measures to block vector-human contact ( e . g . , insecticide spraying ) and vector-dog contact ( e . g . , insecticide impregnated collars and vaccines ) . We suggest that these measures be prioritized for areas that have no current record of vectors or hosts infected with L . i . chagasi , but display invasion potential based on our predictions . Our study represents the first published investigation of the interdependent processes involved in the dispersion of VL competent vectors , infected dogs , and infected humans in São Paulo State and the associated risk factors , showing and anticipating the relentless progressive spread of VL . The prediction of VL dispersion is crucial for the identification of areas of high risk and to enable more spatially targeted prevention and control measures , potentially improving public and animal health policy-making .
|
Visceral leishmaniasis [VL] is an important but neglected tropical disease that occurs worldwide . In areas where the disease is zoonotic , it is considered a serious public and animal health problem . In Brazil , despite the existing prevention and control programs , the disease is spreading , and in São Paulo State its dispersion presents a distinct temporal-geographic pattern . The goal of our study was to understand the landscape , climatic and economic factors that influence the dispersion of visceral leishmaniasis in Sao Paulo State , and then use these findings to predict its spread over time and space . To this end , we integrated data on the sand fly vector of the causative parasitic agent of VL , infected dogs and human cases . We find that landscape and climate were more important than economic factors in predicting competent vector , infected human and infected dog distributions , and the presence of the competent vector and infected dogs strongly influenced the dispersion of infected humans . Our study represents the first integrated investigation of vector and infected host invasion potential for individual municipalities , contributing to VL disease prevention and control planning in São Paulo State .
|
[
"Abstract",
"Introduction",
"Methodology",
"Results",
"Discussion",
"Conclusion"
] |
[
"medicine",
"and",
"health",
"sciences",
"vector-borne",
"diseases",
"tropical",
"diseases",
"vertebrates",
"geographical",
"locations",
"sand",
"flies",
"parasitic",
"diseases",
"dogs",
"animals",
"mammals",
"mathematics",
"neglected",
"tropical",
"diseases",
"infectious",
"disease",
"control",
"insect",
"vectors",
"statistical",
"distributions",
"infectious",
"diseases",
"zoonoses",
"south",
"america",
"epidemiology",
"protozoan",
"infections",
"brazil",
"disease",
"vectors",
"probability",
"theory",
"statistical",
"dispersion",
"people",
"and",
"places",
"leishmaniasis",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"amniotes",
"organisms"
] |
2017
|
Risk analysis and prediction of visceral leishmaniasis dispersion in São Paulo State, Brazil
|
An open question in human genetics is what underlies the tissue-specific manifestation of hereditary diseases , which are caused by genomic aberrations that are present in cells across the human body . Here we analyzed this phenomenon for over 300 hereditary diseases by using comparative network analysis . We created an extensive resource of protein expression and interactions in 16 main human tissues , by integrating recent data of gene and protein expression across tissues with data of protein-protein interactions ( PPIs ) . The resulting tissue interaction networks ( interactomes ) shared a large fraction of their proteins and PPIs , and only a small fraction of them were tissue-specific . Applying this resource to hereditary diseases , we first show that most of the disease-causing genes are widely expressed across tissues , yet , enigmatically , cause disease phenotypes in few tissues only . Upon testing for factors that could lead to tissue-specific vulnerability , we find that disease-causing genes tend to have elevated transcript levels and increased number of tissue-specific PPIs in their disease tissues compared to unaffected tissues . We demonstrate through several examples that these tissue-specific PPIs can highlight disease mechanisms , and thus , owing to their small number , provide a powerful filter for interrogating disease etiologies . As two thirds of the hereditary diseases are associated with these factors , comparative tissue analysis offers a meaningful and efficient framework for enhancing the understanding of the molecular basis of hereditary diseases .
Hereditary diseases arise due to germline aberrations that are present across the human body . Enormous progress has been made over the years in mapping the genetic causes for a large variety of hereditary diseases . To date , causal germline aberrations for over 1 , 500 hereditary diseases can be retrieved from the OMIM database [1] . Additional disease-related factors have been mapped by genome-wide association studies , mRNA profiling and epigenetic marking . Yet , despite this wealth of data , the molecular basis of many hereditary diseases remains elusive . It is thus apparent that novel strategies are required to enhance our understanding of the molecular basis of these diseases . While the genetic aberrations that cause hereditary diseases are global , the diseases are often manifested in specific organs or tissues ( for simplicity we will use ‘tissue’ to denote ‘organ’ as well ) . The mechanism for this selectivity and vulnerability in certain tissues is unknown . Common explanations refer to a unique function of the disease tissue , as in the case of muscle and liver glycogen storage disease , or a unique feature of the disease tissue , such as long-lived neurons and age-related protein misfolding diseases [2] . It was also shown that expression levels of genes underlying genetic diseases tend to be elevated in their disease tissues [3] , [4] . Yet a rigorous analysis of the tissue-specific manifestation of hereditary diseases and their respective disease genes has rarely been performed . An important determinant of tissue features is the repertoire of proteins that are expressed in the tissue and their physical interactions , whose union defines the tissue interactome . Tissue interactomes were utilized to assess tissue-specific functions of proteins and interactions ( e . g . , [4]–[8] ) and to illuminate general properties of disease genes ( e . g . , [3] , [7] ) . Tissue interactomes were typically constructed by superimposing tissue expression data on a static protein-protein interaction ( PPI ) network composed of interactions that were frequently identified through in-vitro assays , without any tissue context . These PPIs were included in the tissue interactome if the interacting partners were found to be expressed in that tissue , otherwise the PPIs were down-weighted or excluded ( e . g . , [4]–[11] ) . Due to the scarcity of protein expression data across tissues , most studies exploited data of gene expression across tissues . The dataset of Su et al . [12] , which used DNA microarrays to profile transcript levels across 77 tissues , has been a prominent quantitative resource in many of these studies ( e . g . , [3] , [9] , [11] ) . Recently , our views of gene and protein expression across tissues were considerably amplified owing to the application of additional technologies to expression mapping . First , application of next-generation RNA sequencing ( RNA-seq ) revealed that many more transcripts were expressed per tissue than previously acknowledged [13] , thereby affecting conclusions drawn from previous data [7] . Second , large-scale measurements of protein abundance have become available , providing direct evidence for the presence of a protein in a tissue . In particular , the Human Protein Atlas ( HPA ) initiative offers qualitative immunohistochemical measurements of protein abundance for thousands of proteins across tens of tissues [14] . Here we took advantage of the exciting new wealth of information regarding gene and protein expression across tissues and harnessed them to illuminate the tissue-specificity of hereditary diseases . First , we combined these data and integrated them with known PPIs to create the interactomes of 16 main human tissues ( Figure 1 ) . We observed considerable similarities between tissue interactomes in terms of expressed proteins and interactions , including significant correlations between transcript levels and the numbers of interacting partners per gene in each tissue . Focusing on genes causing hereditary diseases , we show that , in contrast to the tissue-specific manifestation of their respective diseases , the genes are often expressed in many tissues . However , they tend to have elevated transcript levels or tissue-specific PPIs preferentially in their disease tissues . We demonstrate that knowledge of the tissue-specific PPIs of genes causing hereditary diseases can be used to highlight disease-related mechanisms . Therefore , comparison between tissue interactomes can serve as an efficient strategy for illuminating the molecular basis of these diseases .
We obtained extensive data of genes and proteins expression across tissues by integrating three major datasets: the dataset of the Genomics Institute of the Novartis Research Foundation ( GNF ) measured by using DNA microarrays [12] , the Human Protein Atlas ( HPA ) dataset described above [14] , and the Illumina Body Map 2 . 0 dataset measured by using RNA-seq [15] . From each dataset we extracted the set of proteins or protein-coding genes expressed per tissue , denoted tissue expressome , by using stringent thresholds ( see Methods ) . Since the available PPI data were oblivious to alternatively-spliced isoforms we associated each protein-coding gene with a single protein product ( see Methods ) . For simplicity we henceforth refer to protein-coding genes and proteins interchangeably . Dataset comparison of the genes expressed in each tissue revealed that RNA-seq identified the largest number of genes per tissue , with a median increase of 1 . 5-fold relative to HPA and of 3 . 8-fold relative to GNF ( Table 1 ) . Still , RNA-seq did not fully contain data from the other datasets but covered between 58–80% of their expressomes per tissue . We therefore tested whether the different datasets were compatible and could be combined . Indeed , the overlaps between datasets in expressed genes per tissue were highly statistically significant in all cases ( p-value<10−97 , Fisher exact test ) . Moreover , corresponding tissues from the different datasets best correlated with each other in almost all cases ( see Methods and Table S1 ) . We also compared between the tissue distributions of genes per dataset and found them to be bi-modal , with most genes showing either tissue-specific or ubiquitous expression across tissues ( Figure 2A ) . In particular , in the GNF dataset most genes were tissue-specific , but in the more recent RNA-seq and HPA datasets , in accordance with other expression datasets [4] , most genes were ubiquitously expressed across all tissues . We observed similar bi-modal distributions upon using less stringent expression thresholds ( Figure S1 ) . To obtain an extensive view of the repertoire of expressed genes and their potential PPIs in each tissue we combined the three datasets . Specifically , we defined a gene as expressed in a tissue if it was found to be expressed in that tissue in at least one dataset . The resulting tissue expressomes maintained the bi-modal tissue distribution ( Figure 2A ) : 61% of the genes were expressed in 14–16 tissues , henceforth denoted globally expressed genes , and 14% of the genes were expressed in 1–3 tissues , henceforth denoted tissue-specific genes . Testing for gene ontology ( GO ) enrichments we found that globally expressed genes were highly enriched for basic cellular processes common to living cells , such as RNA splicing ( p<10−28 ) and protein transport ( p<10−26 , Table S2 ) . Tissue-specific genes were enriched for tissue-specific processes such as spermatogenesis in testis ( p = 7 . 3*10−6 ) and sensory perception in brain ( p = 7 . 2*10−4 , Table S3 ) . To construct tissue interactomes we first gathered recent data of experimentally-detected human PPIs from four major public databases [16]–[19] . These data amounted to a global interactome consisting of 67 , 439 interactions between 11 , 225 proteins , thus covering 52% of the human protein-coding genes . We then constructed tissue interactomes by filtering the global interactome according to tissue expressomes ( Figure 1 ) . Specifically , each tissue interactome contained only those PPIs in which both interacting partners were found to be expressed in that tissue ( see Methods ) . Each resulting tissue interactome covered more than half of the global human interactome , with 58–75% of the proteins and 63–79% of the PPIs . The tissue interactomes are provided at http://netbio . bgu . ac . il/tissueinteractoms . An overall view of the tissue interactomes appears in Figure 2B . All the tissue interactomes shared a common core sub-network that contained 4 , 989 proteins and 26 , 370 PPIs . This core sub-network dominated all tissue interactomes by containing half or more of their proteins and PPIs . To test the consistency in expression levels of core proteins across tissues we applied the DESeq method [20] . Only 555 of the 4 , 989 core genes ( 11% ) showed a significant change in expression ( p-value≤0 . 01 ) in at least one tissue , implying that most core proteins are expressed at similar levels across tissues . As can be expected , core proteins were highly enriched for basic cellular processes ( Table S4 ) . Another common feature of the tissue interactomes was the scale-free like distribution of their PPI degrees ( degree is the number of interacting partners per protein ) . In each tissue most proteins had at most five interacting partners , while a small subset of proteins , denoted hubs , had over 45 interacting partners each ( Figure S2 ) . These 451 tissue hubs generally retained their high degree across tissues ( Figure 2C ) , and were highly enriched for a variety of regulatory processes , such as regulation of transcription ( 36% , p<10−15 ) and regulation of signal transduction ( 18% , p<10−12 , see Table S5 ) . 221 of these tissue hubs were also found in the core sub-network and showed a similar regulatory nature relative to other core proteins ( Table S6 ) . The wide range of PPI degrees led us to hypothesize that proteins with many interacting partners may require a larger number of molecules in order to support these interactions , relative to proteins with only a few interacting partners , as was previously observed in budding yeast [21] . We therefore correlated between PPI degrees and transcript level per gene in each tissue ( see Methods ) . In all tissues these correlations were statistically significant ( Figure 2D and Figure S3 ) . These correlations were maintained despite the diversity across tissues in transcript levels and in PPI partners ( Figure S4 ) . We gathered 303 hereditary diseases that manifested clinically in at least one of the 16 tissues that we analyzed , and their 233 causal germline-aberrant genes ( see Methods and Figure S5 ) . As shown in Figure 3A , most hereditary diseases manifested in a single tissue , and yet over 80% of their causal genes were expressed in 10 tissues or more . Thus , causal genes tend to elicit a clear phenotype in only a small subset of their expressing tissues . We next tested whether the expression levels of causal genes differ between their disease- and unaffected tissues , as shown previously for the larger set of genetic diseases caused by somatic or germline aberrations [3] . To this end we compared between the median expression level of causal genes in their disease tissues and their median expression level in non-disease tissues ( see Methods ) . We found that a significant fraction of these genes were expressed at elevated levels in their disease tissues ( 63% , randomization test p<10−4 , Figure 3B ) , with almost a third of these having significantly high levels ( 28% with p≤0 . 01 , DESeq analysis ) . Given the correlation between transcript levels and PPI degrees , we next tested whether causal genes also tend to have more PPIs in their disease tissue . This tendency too was significant ( 42% , randomization test p<10−4 , Figure S6 ) . Moreover , there was a significant overlap between causal genes with elevated expression levels and causal genes with higher PPI degrees in their disease tissue ( Fisher exact test p = 0 . 02 ) . Given that causal genes tended to have more PPIs in their disease tissue , we tested whether they are also associated with PPIs that occur almost exclusively in that tissue . Such tissue-specific PPIs ( TS-PPIs ) can offer a clear molecular basis for the tissue-specific manifestation of hereditary diseases . Indeed , we found several examples where the TS-PPIs of causal genes in their disease tissues involved genes and interactions previously shown to be relevant for disease etiology ( Table 2 and Figure 4 , see Discussion ) . The full list of causal genes and their TS-PPIs appears in Table S8 . We next turned to assess the prevalence of TS-PPIs among causal genes . We found that causal genes had a significantly higher tendency for TS-PPIs relative to interactome genes ( Fisher exact test p = 8 . 8*10−5 , Figure S7 ) . Moreover , their TS-PPIs occurred preferentially in their disease tissues ( randomization test p<10−4 , Figure 3C ) . As could be expected , causal genes with more PPIs and causal genes with TS-PPIs in their disease tissue significantly overlapped ( Fisher exact test p<10−9 ) . However , there was no significant overlap between causal genes with TS-PPIs and causal genes with elevated transcript levels in their disease tissues ( Fisher exact test p = 0 . 56 ) . Thus , TS-PPIs and elevated transcript levels of causal genes in their disease tissues distinctly contribute to the emergence of tissue-specific phenotypes . As shown in Figure 3D , these two factors are related to 67% of the hereditary diseases in our dataset .
The identification of germline-aberrant genes underlying many of the hereditary diseases provides an important step toward unraveling their molecular basis . Still , the remarkable tissue-specificity of hereditary diseases makes it clear that additional factors are governing disease manifestations . Relying on the utility of interactomes in understanding genotype-to-phenotype relationships [22] , we applied here a comparative analysis of tissue interactomes to uncover determinants of the tissue-specificity of hereditary diseases . We analyzed over 300 hereditary diseases and their causal genes . However , this set was limited by several factors: First , only diseases caused by mutations in protein-coding genes were included . Second , diseases had to be associated with at least one of the 16 tissues we analyzed . Given that the tissue associations were deduced by using a text-mining approach [3] , these associations could be noisy or limited to the subset of hereditary diseases with clear tissue phenotypes . Third , the expression level of causal genes in their disease tissue had to reach a certain threshold , and thus diseases whose causal genes are lowly expressed might have been ignored . These limitations imply that our results may be more relevant for monogenic disorders with strong tissue phenotypes . To construct tissue interactomes we combined three large-scale datasets of transcript or protein abundance across a multitude of tissues , which were obtained through three experimental techniques . We found relatively strong correlations between transcript levels in corresponding tissues , and statistically significant yet low correlations between transcript levels and protein abundance . The latter correlation was recently shown to be around 0 . 4 in simultaneous measurements from a common sample [23] . The lower correlations we observed likely stem from noisy estimates of transcript and protein abundance , and from correlating between measurements from different tissue samples . Similarly to other studies of tissue interactomes ( e . g . , [9] , [11] ) , we combined the datasets by associating a gene with a tissue if its expression in that tissue passed certain criteria in at least one dataset . This combination resulted in tissue expressomes that were unique in their extent ( Table 1 ) . At the same time , relying on no more than three sources allowed us to limit lab bias effects that would have been encountered upon analyzing a similar number of samples but from many different labs [24] . We then superimposed the subset of tissue-associated genes on the set of known PPIs , filtering out PPIs in which at least one interacting partner was not associated with the tissue . It is important to note that whether a PPI indeed occurs in the tissue depends on additional factors , such as the cellular localization of the interacting proteins and their posttranslational modifications . Nevertheless , expression of both partners is a necessary initial requirement , and therefore is often used as a filter for constructing tissue interactomes ( e . g . , [3] , [7] , [9]–[11] ) . The effectiveness of filtered tissue interactomes was demonstrated in two recent studies , which showed that they considerably improve the prioritization of disease genes relative to an unfiltered global interactome [10] , [11] . The interactomes of the different tissues had common features . First , the majority of their genes were common to 14 or more interactomes ( Figure 2 ) . These globally expressed genes were enriched in basic cellular processes and formed a common core sub-network that dominated all tissue interactomes ( Figure 3 ) . Second , hubs in the different tissue interactomes were enriched in regulatory processes . Third , the different tissues shared significant correlations between transcript levels and PPI degrees ( Figure 2D and Figure S3 ) . Such correlation was previously observed in budding yeast [21] but not in human . One might assume that these correlations stem from the fact that PPIs between highly-expressed proteins are easier to detect . However , the detection of PPIs was often done outside of a human cell , through in-vitro assays or by using yeast cells ( e . g . [25] ) . In such assays the transcript levels are unrelated to the in-vivo levels of these transcripts within human cells of different tissues , and therefore the bias toward genes with high transcript levels in-vivo is unlikely . In view of these marked similarities between the different tissue interactomes , the hereditary diseases that we analyzed stood out as a critical manifestation of tissue-specificity . Contrary to what might be expected , only 7% of the tissue-specific hereditary diseases were associated with tissue-specific causal genes ( Figure 3D ) . Instead , the large majority of causal genes were expressed in many tissues that , enigmatically , did not show marked disease phenotypes ( Figure 3A ) . We next harnessed the tissue interactomes to identify features that distinguish causal genes in their disease tissues and could underlie the tissue-specific selectivity and vulnerability . We found that causal genes tend to have elevated expression levels in their disease tissues relative to unaffected tissues in which they were expressed ( Figure 3B ) . A similar tendency was observed previously among genes causal for genetic diseases excluding cancers [3] . The correlation we observed between transcript levels and PPI degrees ( Figure 2D ) , and the law of mass-action that links gene dosage with probability of interactions [26] , both suggest that causal genes will interact in a more promiscuous manner in their disease tissues . Indeed , we found that causal genes tend to have more PPIs in their disease tissues . Given that mutations leading to diseases were shown in some cases to disturb the physical interactions of disease proteins [27] , [28] , the higher tendency for potentially disturbed PPIs in disease tissues may underlie the increased vulnerability of these tissues . The other feature that distinguishes causal genes in their disease tissues is their tendency for TS-PPIs , which is observed for 27% of the hereditary diseases ( Figure 3D ) . Notably , such TS-PPIs can offer an explanation for the tissue-specific manifestation of a disease: while they may not comprise the entire disease mechanism , these interactions can enhance or propagate the aberrant phenotypes and thus contribute to the clinical manifestation ( Table 2 ) . An important implication of this observation relates to the interrogation of disease etiologies: Whereas current efforts to illuminate the molecular basis of diseases typically consider all the interactions involving causal genes in their disease tissues ( e . g . , [10] , [11] ) , we suggest concentrating efforts on their TS-PPIs . We show in Table S8 that focusing on these TS-PPIs typically reduces the number of candidate PPIs by 8-fold , thus providing a powerful filter . Below we demonstrate that TS-PPIs can highlight additional disease-related proteins and interactions effectively . Our first example relates to the widely-expressed tumor suppressor BRCA1 that causes predisposition to hereditary breast and ovarian cancer . We found that , in breast , BRCA1 is involved in a single TS-PPI , with the estrogen receptor ESR1 that activates cell proliferation ( Figure 4A ) . Indeed , it was previously demonstrated that through this interaction BRCA1 inhibits ESR1 and its proliferative activity in breast , and that this effect is reduced in mutated forms of BRCA1 [29] . The second example of a disease-related TS-PPI is that of the widely-expressed epidermal growth factor receptor EGFR . Germline and somatic mutations in EGFR lead to lung cancer [30] . Notably , we found that a lung-specific PPI connects EGFR to its ligand-protein epiregulin ( EREG ) that was shown to confer invasive properties in an EGFR-dependent manner [31] . Thus , the EGFR-EREG PPI has the potential to enhance the effect of EGFR aberration in a lung-specific manner ( Figure 4B ) . The third example relates to aberrations in three genes that separately cause different subtypes of muscular dystrophy: dystroglycan 1 ( DAG1 ) , dystrophin ( DMD ) , and caveolin 3 ( CAV3 ) . DMD is a muscle-specific protein that anchors the extracellular matrix to the cytoskeleton . It is also the ligand of DAG1 , a globally expressed trans-membrane cell adhesion receptor that interacts with DMD in muscle only [32] . This interaction is prevented by another muscle-specific PPI , in which CAV3 binds to DAG1 [33] . These muscle-specific PPIs explain the muscle-specific phenotypes of DAG1 and CAV3 aberrations ( Figure 4C ) . The last example is a putative explanation for leukoencephalopathy with vanishing white matter , a brain disease that manifests during childhood . The progressive white matter deterioration is caused by germline mutations in any of the five genes encoding the subunits of the translation initiation factor EIF2B , and cells harboring any of these mutations show decreased translation activity [34] . While EIF2B proteins are globally expressed , we found that they exhibit brain- and testis-specific interactions with the netrin-1-receptor DCC , which mediates axon guidance ( Figure 4D ) . Notably , Tcherkezian et al . [35] that identified the relationships between DCC and EIF2B also showed that the absence of netrin significantly lowers cellular translation . We therefore propose that this relationship enhances the effect of EIF2B mutations in a brain-specific manner . The distinct features we identified provide a starting point for elucidating the molecular basis of many hereditary diseases and can be further applied to filter the wealth of data being generated by large-scale disease-associations studies . In the future , additional tissue features could be considered , such as protein isoform concentrations [36] and relationships other than PPIs [37] . The comparative tissue analysis , along with the extensive resource of human tissue interactomes that we put forward , should become a standard framework for interpreting the wealth of disease-related data and for enhancing our understanding of the etiologies of hereditary diseases .
GNF data [12] were downloaded from BioGPS [38] , and all genes with intensity value above 100 in a tissue were considered as expressed [39] . HPA data [14] included proteins that were identified as expressed in a tissue , i . e . , assigned as ‘low’ , ‘medium’ or ‘high’ abundance based on manual assessment of tissue staining by antibodies against the proteins of interest . Proteins were further filtered by imposing stringent thresholds on the reliability and validity of their antibodies: When available , a medium or high antibody-reliability score was required; otherwise we required at least one supportive and no negative validity scores . In case of multiple measurements per tissue per gene we chose the highest value . RNA-seq data from Illumina Body Map 2 . 0 [15] were filtered for genes with at least 1 read per kilobase per million reads ( RPKM ) . Results for a threshold of 0 . 3 RPKM were similar and appear in Figure S1 . Analysis was limited to proteins and protein-coding genes only , and these were mapped to their Ensembl gene identifiers using BioMart [40] . Table S9 presents the numbers of genes and tissues measured in each dataset . Since RNA-seq data covered the largest number of genes per tissue we based our analysis on the 16 main human tissues profiled with RNA-seq . GNF and HPA each contained profiles for 15 and 14 of these tissues and their subparts , respectively . We manually consolidated the various tissue subparts according to the consolidation scheme given in Table S10 . A gene was considered as expressed in a tissue if that gene or its protein product were found to be expressed in that tissue or the tissue subparts by at least one dataset . Compatibility among datasets was tested by ( i ) computing the overlaps in genes and interactions expressed per tissue using Fisher exact test , and ( ii ) computing the correlation in expression levels of commonly-expressed genes using Kendall's tau rank correlation , and ranking the correlations between matching tissues compared to correlations between non-matching tissues ( Table S1 ) . Figure S8 provides scatter plots comparing the expression levels of common genes in corresponding tissues measured by any two out of HPA , GNF and RNA-seq . We assembled PPI data from BIOGRID [16] , DIP [17] , IntAct [18] and MINT [19] . Only experimentally-detected physical interactions were included , and their union formed the global human interactome . Tissue interactomes were constructed by filtering the PPIs in the global human interactome according to tissue expression data [41] . A PPI was included only if both pair-mates were found to be co-expressed in the same sample or in the same tissue subpart at the lowest hierarchy level ( Table S10 ) . PPIs from subparts of the same tissue were united to form the tissue interactome . Interactome hubs were defined as those nodes in the network where the number of interacting partners ( PPI degree ) ranked among the top 5% . This resulted in a threshold of over 45 PPI partners for each interactome we analyzed . Pair-wise Kendall's tau rank correlations between datasets were computed for expression levels of commonly detected genes . GO enrichments were performed using DAVID [42] . The total number of human proteins was 21 , 450 according to BioMart [40] . Differential expression of genes in a tissue was computed using the DEseq method [20] . DESeq analysis was performed for 16 tissues , such that each run compared one tissue to all other 15 tissues . The statistical significance of the overlap between disease genes with elevated expression , higher PPI degree or TS-PPI was calculated using Fisher exact test while excluding tissue-specific disease genes . In each tissue we computed the Spearman correlation between RPKM levels and PPI degrees of genes with RPKM readout above 0 and PPIs in the tissue . We also binned genes based on their RPKM levels into 10 equally-sized bins , and computed correlations using the median of each bin . In all tissues both types of correlations were highly statistically significant . The correlation values and figures appear in Figure S3 . Disease to tissue associations were taken from Lage et al . [3] , and included manually curated associations or associations exceeding a cutoff of 15 yielding a precision of 85% as mentioned therein . Using data from the OMIM database [1] we limited the set of diseases to include only non-somatic and allelic disorders , and extracted disease-genes that were causally associated with those diseases . Data of hereditary cancers were downloaded from the cancer gene census website [43] , was limited to cancer germline mutation genes , and manually associated with disease tissues . Only diseases that were associated with at least one of the 16 main tissues and whose causal disease gene was expressed in that tissue were analyzed . We compared the RPKM values of a disease gene between its disease and non-disease tissues in two ways . To identify elevated expression we compared between ( i ) the median RPKM of that gene in its disease tissues , and ( ii ) the median RPKM of that gene in its non-disease tissues . The permutation test used to assess the significance of the results is described in the following subsection . To identify significant over-expression we used the DEseq method as described in the ‘statistical analysis’ subsection above [20] . The subset of tissues considered for a specific gene was limited to tissues in which the gene was indeed expressed . We used a permutation test to assess the statistical significance of the number of genes whose median value ( RPKM level , PPI degree or number of TS-PPI ) in their disease tissues was higher than the corresponding median value in their non-disease tissues . Specifically , for each disease gene we randomly selected a set of x disease tissues out of the set of tissues expressing that gene , where x was set to the number of original disease tissues for that gene . The remaining , non-selected tissues expressing the gene were considered as the gene's random non-disease tissues . In all calculations we ignored tissue-specific disease genes as their set of non-disease tissues could be empty . We then computed the relevant median value for that gene in its randomly-selected disease tissues ( median value denoted V_d ) and the median value in its random non-disease tissues ( median value denoted V_nd ) . If the median in disease tissues was higher ( i . e . , V_d>V_nd ) the gene was considered as success in the permutation test . We applied this test to all disease genes and counted the total number of random successes in that run . We repeated this analysis 10 , 000 times . We computed the p-value as the fraction of runs out of the 10 , 000 runs in which the total number of random successes was at least as high as the number of successes in the original data . The supplementary material file contains Figures S1 , S2 , S3 , S4 , S5 , S6 , S7 , S8 and Tables S1 , S2 , S3 , S4 , S5 , S6 , S7 , S8 , S9 , S10 , S11 , S12 , S13 . The tissue interactomes can be found at http://netbio . bgu . ac . il/tissueinteractoms .
|
An open question in human genetics is what underlies the tissue-specific manifestation of hereditary diseases , which are caused by genomic aberrations that are present in cells across the entire human body . In order to answer this question , we created an extensive resource of protein expression and interactions across 16 main human tissues . Using this resource , we first show that the genes underlying hundreds of hereditary diseases are widely expressed across tissues , yet , enigmatically , cause disease phenotypes in few tissues only . We then identify two distinct , statistically-significant factors that could lead to tissue-specific vulnerability in the face of this broad expression: ( i ) many disease-causing genes have elevated expression levels in their disease tissues , and ( ii ) disease-causing genes have a significantly higher tendency for tissue-specific interactions in their disease tissues . As we show for several disease-causing genes , these tissue-specific interactions highlight disease mechanisms and provide an efficient filter for interrogating the molecular basis of diseases . Together the two factors we identified are relevant for as many as two thirds of the tissue-specific hereditary diseases . Our comparative tissue analysis therefore provides a meaningful and efficient framework for enhancing the understanding of the molecular basis of hereditary diseases .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"systems",
"biology",
"biochemistry",
"genetic",
"networks",
"genomics",
"functional",
"genomics",
"protein",
"interactions",
"proteins",
"genome",
"analysis",
"genetics",
"biology",
"and",
"life",
"sciences",
"proteomics",
"computational",
"biology"
] |
2014
|
Comparative Analysis of Human Tissue Interactomes Reveals Factors Leading to Tissue-Specific Manifestation of Hereditary Diseases
|
Reptile-associated Salmonella bacteria are a major , but often neglected cause of both gastrointestinal and bloodstream infection in humans globally . The diversity of Salmonella enterica has not yet been determined in venomous snakes , however other ectothermic animals have been reported to carry a broad range of Salmonella bacteria . We investigated the prevalence and diversity of Salmonella in a collection of venomous snakes and non-venomous reptiles . We used a combination of selective enrichment techniques to establish a unique dataset of reptilian isolates to study Salmonella enterica species-level evolution and ecology and used whole-genome sequencing to investigate the relatedness of phylogenetic groups . We observed that 91% of venomous snakes carried Salmonella , and found that a diverse range of serovars ( n = 58 ) were carried by reptiles . The Salmonella serovars belonged to four of the six Salmonella enterica subspecies: diarizonae , enterica , houtanae and salamae . Subspecies enterica isolates were distributed among two distinct phylogenetic clusters , previously described as clade A ( 52% ) and clade B ( 48% ) . We identified metabolic differences between S . diarizonae , S . enterica clade A and clade B involving growth on lactose , tartaric acid , dulcitol , myo-inositol and allantoin . We present the first whole genome-based comparative study of the Salmonella bacteria that colonise venomous and non-venomous reptiles and shed new light on Salmonella evolution . Venomous snakes examined in this study carried a broad range of Salmonella , including serovars which have been associated with disease in humans such as S . Enteritidis . The findings raise the possibility that venomous snakes could be a reservoir for Salmonella serovars associated with human salmonellosis .
Salmonella is a clinically relevant bacterial pathogen that poses a significant burden upon public health worldwide [1–4] . Two groups of Salmonella serovars have clinical relevance with distinct host-specificity and disease manifestations . Typhoidal Salmonella is restricted to human hosts and presents as a systemic infection , resulting in an estimated 223 , 000 fatalities per annum [5] . In contrast , nontyphoidal Salmonella typically manifests as a self-limiting gastrointestinal disease in otherwise healthy individuals around the world , causing an annual global disease burden of 93 . 8 million cases and 155 , 000 deaths [3] . Over the past two decades , an invasive form of nontyphoidal Salmonella ( iNTS ) has emerged as the most prevalent bacterial species to be isolated from the bloodstream of patients in sub-Saharan Africa [6] . Over 3 . 4 million cases and 680 , 000 deaths are estimated to occur worldwide each year as a result of iNTS [4] . The Salmonella genus contains two species; S . bongori and S . enterica . S . enterica is further divided into six subspecies; enterica ( I ) , salamae ( II ) , arizonae ( IIIa ) , diarizonae ( IIIb ) , houtanae ( IV ) and indica ( VI ) [7] . The subspecies are classified into approximately 2 600 serovars which are ecologically , phenotypically and genetically diverse [8] . Serovars which belong to S . enterica subspecies enterica cluster phylogenetically into two predominant clades ( A and B ) [9–11] . Here , we use the term Salmonella to refer to the S . enterica species and the S . enterica designation to refer to the S . enterica subspecies enterica alone , unless stated otherwise . Biochemical properties such as carbon utilisation and anaerobic metabolism are often serovar-specific [12] . The ability of Salmonella to grow in a wide range of conditions reflects the adaptation of the bacteria to survive in the environment or in different hosts , as demonstrated by a recent study focused on genome-scale metabolic models for 410 Salmonella isolates spanning 64 serovars in 530 different growth conditions [13] . At the genus level , Salmonella has a broad host-range whilst individual serovars differ in host-specificity [14] . The majority of Salmonella infections in humans ( 99% ) are caused by a small number of serovars belonging to the S . enterica subspecies [15] . Serovars which belong to non-enterica subspecies are associated with carriage in ectothermic animals such as reptiles and amphibians , but are rarely found in humans [14 , 16–18] . Carriage rates of non-enterica serovars in reptiles can be high . A study focused on snakes in a pet shop found that 81% of animals were carrying S . diarizonae [18] . Previous studies demonstrating the diverse range of Salmonella subspecies that colonise various reptilian species in different countries are summarised in Table 1 . Reptiles represent a significant reservoir for serovars of Salmonella that are associated with human disease . Over 60% of captive-bred reptiles between 1995 and 2006 in Denmark were reported to carry S . enterica subspecies enterica serovars [24] . About 6% of human salmonellosis cases were contracted from reptiles in the USA [25] , and in South West England , 27 . 4% of Salmonella cases in children under five years old were linked to reptile exposure [26] . The latter study demonstrated that reptile-derived salmonellosis was more likely to cause bloodstream infection in humans than non-reptile-derived Salmonella [26] . Reptile-associated Salmonella is therefore considered to be a global threat to public health [27] . The majority of reptile-associated salmonellosis cases reported in humans are caused by Salmonella from non-venomous reptiles [27] , probably because these animals are frequently kept as pets . Therefore , non-venomous reptiles have been the focus of numerous studies whilst the prevalence and diversity of Salmonella in venomous snakes has remained unknown . The recent inclusion of snakebite as a neglected tropical disease demonstrates that these reptiles frequently interact with humans in tropical and sub-tropical countries . The proximity of venomous snakes to humans may lead to contaminated faecal matter being shed on the surfaces and in water sources used for human homes and to irrigate salad crops [28–30] . Research to improve snakebite treatment at the Liverpool School of Tropical Medicine ( LSTM ) has resulted in the creation of the most extensive collection of venomous snakes in the UK ( 195 ) . The LSTM herpetarium houses venomous snakes from a diverse range of species and geographical origins , representing an ideal source of samples to assess Salmonella in this under-studied group of reptiles . The aims of this study were three-fold . Firstly , to determine the period prevalence of Salmonella in a collection of captive venomous snakes and investigate whether this group of reptiles could act as reservoirs for human salmonellosis . Secondly , to assess the serological and phylogenetic diversity of Salmonella amongst reptiles . Thirdly , to use the diversity of reptile-associated Salmonella to determine clade-specific differences that could reflect adaptation to survival in the environment or to different hosts . Here , we present the first whole genome-based comparative study of the Salmonella bacteria that colonise venomous and non-venomous reptiles .
The Salmonella isolates were derived from faecal samples from two collections of reptiles . One hundred and six faecal samples were collected from venomous snakes at LSTM from May 2015 to January 2017 , with an emphasis on snakes originating from Africa ( S1 Table ) , and investigated for the presence of Salmonella . All venomous snakes were housed in individual enclosures and fed with frozen mice . Sixty-nine of the samples ( 71% ) were sourced from wild-caught snakes originating from: Togo , Nigeria , Cameroon , Egypt , Tanzania , Kenya , South Africa , and Uganda . A further 28 Salmonella isolates ( 29% ) came from venomous snakes bred in captivity . The LSTM herpetarium is a UK Home Office licensed and inspected animal holding facility . A second collection of 27 Salmonella isolates from non-venomous reptiles and 1 Salmonella isolate from a venomous reptile were sourced from the veterinary diagnostics laboratory based at the University of Liverpool’s Leahurst campus ( reptilian species described in S1 Table ) . These isolates were collected from June 2011 to July 2016 from specimens submitted as part of Salmonella surveillance for import/export , in addition to veterinary faecal samples and tissues from post mortem investigations . The provenance of the isolates is described in S1 Table . The majority of the non-venomous reptiles were sourced from a zoological collection , however two animals were privately owned and three were sourced from the Royal Society for the Prevention of Cruelty to Animals ( RSPCA ) . The LSTM isolates are henceforth referred to as venomous snake isolates and the Leahurst isolates are referred to as non-venomous reptile isolates unless otherwise stated . All media were prepared and used in accordance with the manufacturer’s guidelines unless otherwise stated . Salmonella was isolated using a modified version of the protocol described in the national Standard Operating Procedure for detection of Salmonella issued by Public Health England [31] . Faecal droppings were collected from reptiles and stored in 15 mL plastic centrifuge tubes at 4°C . Two different methods were used for the enrichment of Salmonella from faecal samples due to reagent availability at the time of isolation . S1 Table provides information on isolate specific methods . In enrichment method 1 , faecal samples were added to 10 mL of buffered peptone water ( Fluka Analytical , UK , 08105-500G-F ) and incubated overnight at 37°C with shaking at 220 rpm . Following overnight incubation , 100 μL of the faeces mixture was added to 10 mL of Selenite Broth ( 19 g/L Selenite Broth Base , Merck , UK , 70153-500G and 4 g/L Sodium Hydrogen Selenite , Merck 1 . 06340-50G ) and incubated overnight at 37°C with shaking at 220 rpm . In enrichment method 2 , faecal samples were added to 10 mL of Buffered Peptone Water ( Fluka Analytical , 08105-500G-F ) supplemented with 10 μg/mL Novobiocin ( Merck , N1628 ) , and incubated overnight at 37°C with shaking at 220 rpm . Following overnight incubation , 100 μL of the faeces mixture was added to 10 mL of Rappaport-Vassilliadis Medium ( Lab M , UK , LAB086 ) and incubated for 24 hours at 42°C with shaking at 220 rpm . Following enrichment , 10 μL of overnight broth was spread onto Xylose Lysine Deoxycholate ( XLD ) ( Oxoid , UK , CM0469 ) agar plates which were incubated overnight at 37°C . Putative Salmonella colonies were selected by black appearance on XLD plates and confirmed by pink and white colony formation on Brilliant Green Agar ( Merck , 70134-500G ) supplemented with 0 . 35 g/L Mandelic Acid ( Merck , M2101 ) and 1 g/L Sodium Sulfacetamide ( Merck , S8647 ) . To identify S . enterica species , colony PCR of the Salmonella specific ttr locus , which is required for tetrathionate respiration [32] , was performed . PCR reagents included MyTaq Red Mix 1x ( Bioline , UK , BIO-25043 ) , ttr-4 reverse primer ( 5'-AGCTCAGACCAAAAGTGACCATC-3' ) and ttr-6 forward primer ( 5'-CTCACCAGGAGATTACAACATGG-3' ) on colonies suspected to be Salmonella . PCR reaction conditions were as follows: 95°C 2 min , 35 x ( 95°C 15 s , 60°C 30 s , 72°C 10 s ) , 72°C 5 min . PCR products were visualised using agarose gel ( 3 . 5% ) ( Bioline , BIO-41025 ) electrophoresis in TAE buffer . Midori Green DNA stain ( 3 μL/100 mL ) ( Nippon Genetics , Germany , MG 04 ) was used to visualise DNA bands under UV light . Throughout the isolation procedure , S . enterica serovar Typhimurium ( S . Typhimurium ) strain LT2 [33 , 34] was used as a positive control , and Escherichia coli MG1655 [35] was used as a negative control ( S1 Table ) . All non-venomous reptile isolates , one venomous reptile isolate and 87 of 97 venomous snake isolates were sent for whole-genome sequencing . Isolates were sent to either MicrobesNG , UK or the Earlham Institute , UK for whole-genome sequencing on the Illumina HiSeq platform ( Illumina , California , USA ) . Isolates which were sequenced by MicrobesNG were prepared for sequencing in accordance with the company’s preparation protocol for single colony-derived bacterial cultures ( http://www . microbesng . uk ) . Isolates which were sequenced by the Earlham Institute were prepared by inoculating a single colony of Salmonella into a FluidX 2D Sequencing Tube ( FluidX Ltd , UK ) containing 100 μL of Lysogeny Broth ( LB , Lennox ) and incubating overnight at 37°C , with shaking at 220 rpm . LB was made using 10 g/L Bacto Tryptone ( BD Biosciences , UK , 211705 ) , 5 g/L Bacto Yeast Extract ( BD , 212750 ) and 5 g/L Sodium Chloride ( Merck , S3014-1kg ) . Following overnight growth , the FluidX 2D Tubes were placed in a 95°C oven for 20 minutes to heat-kill the isolates . DNA extractions and Illumina library preparations were conducted using automated robots at MicrobesNG or the Earlham Institute . At the Earlham Institute , the Illumina Nextera XT DNA Library Prep Kit ( Illumina , FC-131-1096 ) was used for library preparation . High throughput sequencing was performed using an Illumina HiSeq 4000 sequencing machine to generate 150 bp paired-end reads . Sequencing was multiplexed with 768 unique barcode combinations per sequencing lane . The insert size was approximately 180 bp , and the median depth of coverage was 30x . At MicrobesNG ( https://microbesng . uk ) , genomic DNA libraries were prepared using the Nextera XT Library Prep Kit ( Illumina , FC-131-1096 ) with two nanograms of DNA used as input and double the elongation time that was described by the manufacturer . Libraries were sequenced on the Illumina HiSeq 2500 using a 250 bp protocol . The Salmonella In Silico Typing Resource ( SISTR ) v1 . 0 . 2 was used for serovar prediction [36] . Enterobase [37] was used to assign a Multi Locus Sequence Type ( MLST ) to each isolate , based on sequence conservation of seven housekeeping genes [37] . Where available , reference isolates representing previously sequenced Salmonella isolates for all subspecies and serovars identified were included in the analysis . Reference sequence assemblies were downloaded from the National Center for Biotechnology Information ( NCBI ) . Accession numbers are available in S2 Table . Fastqc v0 . 11 . 5 ( https://www . bioinformatics . babraham . ac . uk/projects/fastqc/ ) and multiqc v1 . 0 ( http://multiqc . info ) were used to assess read quality . Kraken v0 . 10 . 5-beta [38] was run to ensure reads were free from contamination using the MiniKraken 8gb database and a Salmonella abundance cut-off of 70% . Trimmomatic v0 . 36 [39] was then used on the paired end reads to trim low-quality regions using a sliding widow of 4:200 . ILLUMINACLIP was used to remove adapter sequences . Genomes were assembled using SPADES v3 . 90 [40] . QUAST v4 . 6 . 3 [41] was used to assess the quality of assemblies , the results of which can be found in S3 Table . Assemblies which comprised of greater than 500 contiguous sequences were deemed too fragmented for downstream analysis . All assemblies which passed QC were annotated using Prokka v1 . 12 [42] . Roary v3 . 11 . 0 [43] was used to generate a core genome alignment . SNP-sites v2 . 3 . 3 [44] was used to extract SNPs . A maximum likelihood tree was built from the core genome SNP alignment of all isolates using RAxML-NG v0 . 4 . 1 BETA [45] with the general time reversible GTR model and gamma distribution for site specific variation and 100 bootstrap replicates to assess support . The tree was rooted using the Salmonella species S . bongori . Interactive Tree Of Life v4 . 2 [46] was used for tree visualisation . We confirmed that there was no bias in phylogenetic signal between the two different sequencing platforms used by assessing clustering patterns within the phylogenies . S1 Table contains details of the sequencing facility . Monophyletic clustering of isolates was used to assign subspecies to newly sequenced Salmonella from venomous and non-venomous reptilian hosts . The level of association between venom status and phylogenetic clade was determined using odds ratios and χ2 statistics using the OpenEpi website ( http://www . openepi . com ) . Genes involved in the utilisation of each carbon source were identified using KEGG [47] and relevant literature [9 , 32 , 48–59] ( see S4 Table ) . Genes involved in the uptake of carbon sources were prioritised . Sequences were downloaded using the online tool SalComMac [60] , which allows the download of fasta sequences of the genes in S . Typhimurium strain 4/74 . In the case of lac genes , the sequences were taken from the E . coli reference sequence MG1655 . The sequences can be found in S1 Text . The software tool MEGABLAST v2 . 2 . 17 [61] was used to perform a BLAST search of genes in the reptile-derived genomes against a custom-made database of genes diagnostic of Salmonella Pathogenicity Islands and genes involved in carbon utilisation . To confirm all MEGABLAST results , the short reads were mapped against each gene using BWA v0 . 7 . 10 [62] and SAMtools v0 . 1 . 19 [63] . The resulting bam files were manually assessed for gene presence and absence using Integrative Genomics Viewer v2 . 4 . 15 [64] . The results were plotted against the maximum likelihood phylogeny using Interactive Tree Of Life v4 . 2 [46] . Differential carbon source utilisation of 39 reptile-derived Salmonella isolates from S . diarizonae , S . enterica clade A and S . enterica clade B was assessed . Filter-sterilised carbon sugar solutions were added into M9 ( Merck , M6030-1kg ) agar at concentrations detailed in S5 Table . Isolated colonies were transferred from LB agar plates onto M9 carbon source plates using a sterile 48-pronged replica plate stamp and incubated at 37°C under aerobic conditions . An LB control plate was used to validate successful bacterial transfer and all experiments were performed in duplicate . If no growth was seen under aerobic conditions for a particular carbon source , the procedure was repeated under anaerobic conditions ( approx . 0 . 35% oxygen ) with 20 mM Trimethylamine N-oxide dihydrate ( TMAO ) ( Merck , 92277 ) as a terminal electron acceptor . Anaerobic conditions were achieved by incubating plates in an anaerobic jar with 3x AnaeroGen 2 . 5 L sachets ( Thermo Scientific , UK , AN0025A ) to generate anaerobic gas . Oxygen levels were measured using SP-PSt7-NAU Sensor Spots and the Microx 4 oxygen detection system ( PreSens , Regensburg , Germany ) . Salmonella growth was determined at 18 , 90 and 162 hours in aerobic growth conditions and at 162 hours in anaerobic growth conditions . A sub-set of growth positive isolates were assessed for single colony formation to validate the results of the replica plating . Antimicrobial susceptibility was determined using a modified version of the European Committee on Antimicrobial Susceptibility Testing ( EUCAST ) disk diffusion method [65] using Mueller Hinton ( Lab M , LAB039 ) agar plates and a DISKMASTER dispenser ( Mast Group , UK , MDD64 ) . Inhibition zone diameters were measured and compared to EUCAST zone diameter breakpoints for Enterobacteriaceae [66] . Isolates were first tested with six commonly used antibiotics ( Ampicillin 10 μg , Chloramphenicol 30 μg , Nalidixic Acid 30 μg , Tetracycline 30 μg , Ceftriaxone 30 μg , and Trimethoprim/Sulfamethoxazole 25 μg ) and were then tested with five additional antibiotics ( Meropenem 10 μg , Gentamicin 10 μg , Amoxicillin/Clavulanic Acid 30 μg , Azithromycin 15 μg , and Ciprofloxacin 5 μg ) if any resistance was seen to the primary antibiotics ( all disks from Mast Group ) . If resistance was observed phenotypically , then the presence of antimicrobial resistance genes were investigated using the Resfinder software tool [67] . Antimicrobial resistance was defined as resistance to one antimicrobial to which isolates would normally be susceptible [68] . Multidrug resistance was defined as an isolate which showed resistance to three or more antimicrobials to which it would normally be susceptible [68] .
Salmonella carriage is well documented amongst reptiles ( Table 1 ) , however , to our knowledge no published study reports the incidence of Salmonella in venomous snakes . The period prevalence of Salmonella was assessed in a collection of 106 venomous snakes housed at the LSTM venom unit between May 2015 and January 2017 . A remarkably high proportion ( 91%; 97/106 ) of the faecal samples contained Salmonella ( S1 Table ) , which should be seen in the context of the significant carriage rate of Salmonella by other non-venomous reptiles described in the literature [27] . Variable rates of Salmonella carriage have been observed in collections of reptiles ( Table 1 ) , and the large proportion of venomous snakes carrying Salmonella in our study sits at the higher end of the reported spectrum . Our findings pose important public health considerations for individuals who work with venomous snakes housed in captivity , which may previously have been overlooked . To assess diversity , 87 venomous snake-derived Salmonella isolates , 27 non-venomous reptile-derived Salmonella isolates and one venomous reptile-derived Salmonella isolate were whole-genome sequenced . In silico serotyping revealed 58 different Salmonella serovars ( Fig 1 ) . A wide range of serovars was found in each of the venomous and non-venomous snake collections . Given that many of the serovars identified have a very broad host specificity , we suggest that the presence of particular serovars is not linked to the venom status of the reptile . Similar levels of Salmonella carriage were seen in wild-caught and captive-bred reptiles . It is likely that the difference in serovar distribution reflects the sourcing of the reptiles from two independent housing facilities , and represents a limitation of our study . Nevertheless , an unprecedented level of Salmonella diversity was identified amongst both venomous and non-venomous reptiles . Following whole-genome sequencing , multi-locus sequence typing ( MLST ) was used for sub-serovar genetic characterisation of Salmonella [37] . In all cases , isolates falling within the same serovar had identical sequence types ( S1 Table ) , reflecting the intra-serovar homogeneity of the Salmonella isolated in this study . The most common serovar to be identified amongst the venomous snake isolates was S . Souhanina ( n = 12 ) . All S . Souhanina isolates clustered locally on the phylogeny ( Fig 2 ) falling within a 5 SNP cluster characteristic of a clonal expansion event [70] . Four of these isolates were found in captive-bred reptiles , whilst 11 isolates came from venomous snakes which originated in Cameroon , Uganda , Tanzania , Nigeria , Togo and Egypt . The close phylogenetic relationship between the S . Souhanina isolates that belong to the same MLST type ( ST488 ) from imported animals with a range of origins and from captive animals suggests that local Salmonella transmission may be occurring . Local transmission of near-identical salmonellae could occur between snakes or as a result of a single contaminated food source such as frozen mice [71 , 72] . Although our data suggest that S . Souhanina was locally transmitted within the herpetarium , a single source of Salmonella would not explain the wide variety of serovars and MLST types carried by this collection of venomous snakes . Significant Salmonella diversity was reported in a study that involved 166 faecal samples from wild-caught reptiles in Spain , identifying 27 unique serovars [73] . An assessment of Salmonella diversity in wildlife in New South Wales , Australia identified 20 unique serovars amongst 60 wild-reptiles [74] . We speculate that the majority of the diversity of Salmonella identified here originated from wild-caught reptiles and reflect their varied habitats . Underpinning our strategy for sampling Salmonella was the assumption that venomous snakes can carry and shed Salmonella for long periods of time . The longitudinal shedding of Salmonella has been reported in 12 captive non-venomous snakes from 7 different species . Over 10 consecutive weeks , 58% of the snakes shed Salmonella intermittently [75] . Chronic Salmonella carriage has been reported in many other animals , including laying hens which shed the bacteria continually for up to 10 weeks [76] . To assess the continuity of Salmonella shedding from venomous snakes in this study we collected three faecal samples from a Western Green Mamba from Togo over a three-month period between 31st October 2016 and 31st January 2017 . All three faecal samples contained Salmonella which belonged to sequence type ST488 , showing that individual snakes have the capacity to shed the same sequence type of Salmonella over a 90-day period in this study . We propose that the majority of the reptile-derived Salmonella described in this study were acquired by reptiles prior to captivity , whilst some isolates were transmitted locally within the herpetarium . Because the majority of venomous snakes examined in this study were of African origin or belonged to a species of snake native to the African continent ( Fig 2 ) , we compared the Salmonella serovars isolated from all reptiles in this study with those most frequently associated with human disease in Africa . The Salmonella serovar distribution has been reported by the WHO global foodborne infections network data bank based on data from quality assured laboratories in Cameroon , Senegal and Tunisia [69] ( Fig 1 ) . Eleven snake-derived isolates belonged to serovars commonly pathogenic in humans . This finding prompted us to determine the proportion of all venomous snakes and non-venomous reptiles that carried antimicrobial resistant Salmonella ( Table 2 ) . In Salmonella collected from venomous snakes , 4 . 1% of isolates ( 4/97 ) were resistant to at least one antimicrobial and two isolates were multidrug resistant ( Table 2 ) . Three resistant isolates from venomous snakes belonged to the serovar Enteritidis and were closely related to the global S . Enteritidis epidemic clade which causes human disease in Africa [77] . These findings demonstrate that venomous snakes are capable of carrying and shedding Salmonella that have the potential to cause disease in humans . Here we have shown that venomous snakes can shed Salmonella . The vast diversity of Salmonella has long been acknowledged in the literature [37] . To study the diversity of reptile-associated Salmonella from an evolutionary perspective , we obtained 87 high quality whole-genome sequences for a phylogenetic comparison that involved 24 contextual Salmonella genomes ( methods ) . The 87 genomes represented 60 isolates from venomous snakes , 26 Salmonella isolates from non-venomous reptiles and 1 Salmonella isolate from a venomous reptile . Following a comprehensive comparative genomic analysis , we identified a total of 405 , 231 core genome SNPs that differentiated the 87 isolates , and were used to infer a maximum likelihood phylogeny ( Fig 2 ) . SNPs are a valuable marker of genetic diversity [70] , and the identification of hundreds of thousands of core-genome SNPs reflects a high level of genetic diversity among the reptile associated Salmonella isolates . The collection of reptile-derived Salmonella represented most of the known diversity of the Salmonella genus [7] , spanning four of the six Salmonella enterica subspecies: diarizonae , enterica , houtanae and salamae . Reptile-derived S . enterica subspecies enterica isolates were approximately equally distributed among two distinct phylogenetic clusters , known as clade A ( 58% ) and clade B ( 48% ) [9–11] ( Fig 2 ) . No significant association was found between venom status and phylogenetic group ( OR = 1 . 1 , CI = 0 . 3–3 . 0 , χ2 = 0 . 02 , P = 0 . 4 ) . The unique collection of diverse Salmonella isolates was used to determine the phenotypic and genotypic conservation of infection-relevant properties and genomic elements . Whilst the reptile-associated Salmonella belonged to five evolutionary groups , the majority of isolates were classified as S . diarizonae or S . enterica . The clustering of S . enterica into two clades ( A and B ) has previously been inferred phylogenetically based on the alignment of 92 core loci [9 , 10] . The biological significance of S . enterica clade A and clade B has been established as the two clades differ in host specificity , virulence-associated genes and metabolic properties such as carbon utilisation [11] . The genome sequences were used to expand upon pre-existing knowledge and determine phenotypic and genotypic conservation of metabolic and virulence factors across S . diarizonae and S . enterica ( clades A and B ) . Although the majority of Salmonella serovars of public health significance belong to clade A , certain clade B serovars such as Salmonella Panama have been associated with invasive disease [78 , 79] . The clade B S . enterica generally carry a combination of two Salmonella genomic islands . The Salmonella Pathogenicity Island-18 encodes an intracellularly expressed pore forming hemolysin hlyE and the cytolethal distending toxin islet which includes the gene cdtB [2 , 9] . It has been suggested that the two islands are associated with invasive disease , as previously they had only been identified in S . enterica serovar Typhi and Paratyphi A , which cause bloodstream infections [2 , 9] . The combination of hlyE and cdtB genes were present in all S . diarizonae and S . enterica clade B isolates in this study , but absent from all but one S . enterica clade A isolate ( 14L-2174 ) . We propose that the significant proportion of reptiles which carried S . enterica clade B could partially explain the increased likelihood of reptile-associated salmonellosis involving invasive disease , compared to non-reptile-acquired salmonellosis [26] . To assess metabolic differences that distinguish S . enterica clade A , S . enterica clade B and S . diarizonae , we phenotypically screened 39 reptile isolates for the ability to catabolise a number of infection-relevant carbon sources [48 , 77 , 80 , 81] ( S4 Table and S5 Table ) . A summary of the results for phenotypic carbon utilisation and the presence of genes associated with the cognate metabolic pathway is shown in Fig 3 . In general , the genotype accurately reflected phenotype in terms of carbon source utilisation; however , this was not always the case ( Fig 3 ) . Discrepancies between phenotypic growth and genotype suggests that mechanisms of Salmonella metabolism remain to be elucidated . For example , S . diarizonae isolate LSS-18 grew well on myo-inositol as a sole carbon source ( Fig 3 ) but showed zero percent homology with any of the iol genes from the well-characterised Salmonella strain 4/74 . The ability to utilise lactose was a property of most S . diarizonae isolates , consistent with previous reports that 85% of S . diarizonae are Lac+ [82] . It is estimated that less than 1% of all Salmonella ferment lactose due to the loss of the lac operon from the S . enterica subspecies [83] . It was interesting to discover that one non-venomous snake isolate ( 13L-2837 ) which belongs to S . enterica clade B was capable of utilising lactose as a sole carbon source . Isolate 13L-2337 belongs to the serovar S . Johannesburg and to our knowledge this is the first published occurrence of a Lac+ S . Johannesburg isolate . The 13L-2837 pan-genome had zero percent homology to the lac genes from reference strain E . coli MG1655 ( sequence in S1 Text ) ( results in Fig 3 ) , suggesting an alternative method for lactose utilisation . The 13L-2837 S . Johannesburg isolate also lacked the ability to grow on dulcitol , despite possessing all of the relevant gat genes , raising the possibility of an inverse relationship between the ability of Salmonella to utilise dulcitol and lactose as a sole carbon source . These findings require further investigation which is beyond the scope of the current study . The majority of S . enterica clade A and clade B isolates utilised dulcitol , whereas dulcitol was rarely used as a sole carbon source by S . diarizonae . These findings are consistent with a study of Salmonella derived from Australian sleepy lizards , which demonstrated that dulcitol utilisation was observed in almost all S . enterica and S . salamae isolates but only 10% of S . diarizonae isolates [84] . Over 50% of the S . enterica clade A isolates lacked the gatY gene but grew well on dulcitol as a sole carbon source , suggesting that GatY is not required for dulcitol catabolism . A variety of repertoires of dulcitol catabolic genes have been described across Salmonella , with individual serovars carrying one of two gat gene clusters [59] . Both of these clusters carry the gatY gene . Our findings may indicate that a third gat gene cluster is carried by some Salmonella serovars . In the majority of cases , allantoin was only utilised as a sole carbon source by S . enterica clade A isolates , consistent with a previous report that described an association of clade A with the allantoin catabolism island [9] . The majority of clade B isolates lacked the allantoin catabolism island and thus were unable to utilise allantoin as a sole carbon source . However , we identified one clade B isolate as an exception , isolate 11L-2351 which was sampled from a non-venomous reptile . This isolate belongs to the serovar Montevideo , which is frequently associated with outbreaks of human salmonellosis [85–87] . In reptiles , the end product of the purine catabolic pathway is not allantoin , but uric acid [9] . The consequent absence of allantoin from the snake gastrointestinal tract could explain why a substantial number of S . enterica clade B were found in snakes . It is possible that the gain and loss of allantoin catabolic genes is relevant to host specificity . A relationship between the pseudogenization of the allantoin metabolic genes and niche adaptation has also been proposed for the invasive nontyphoidal Salmonella ( iNTS ) reference isolate for S . Typhimurium: D23580 [49 , 88] . Compared with S . Typhimurium isolate 4/74 , which shows a broad host range , D23580 is unable to utilise allantoin as a sole carbon source , consistent with the adaptation of invasive Salmonella in Africa towards non-allantoin producing hosts [49 , 88] . Furthermore , accumulation of pseudogenes in the allantoin degradation pathway has been reported in host-restricted Salmonella serovars which cause invasive disease , suggesting that the ability to grow on allantoin is a marker of a switch from enteric to invasive disease [89] . These findings may reflect the clinical observation that snake-acquired salmonellosis is frequently an invasive disease that commonly results in hospitalisation , compared to disease caused by Salmonella derived from allantoin-producing hosts . Although reptiles are known to harbour a diverse range of Salmonella bacteria , until now Salmonella carriage has not been examined in many key reptilian species . Here , we have shown that venomous snakes harbour and shed a wide variety of Salmonella serovars that represent much of the spectrum of the Salmonella genus and are phylogenetically distributed in a similar way to Salmonella found in non-venomous reptiles . We demonstrated that venomous snakes can carry and excrete Salmonella serovars which cause human disease . One of the Salmonella isolates was resistant to first-line antimicrobial agents . It is possible that venomous snakes represent a previously uncharacterised reservoir for Salmonella both in captive settings and in the wider environment . Further study is required to investigate the relationship between clinical cases and reptile-derived Salmonella in tropical regions inhabited by venomous reptiles such as Africa . We believe that our study provides a good baseline for this future work . Reptiles are an ideal population of animals for the study of genus-level evolution of Salmonella because they carry phylogenetically diverse isolates that belong to the majority of Salmonella subspecies . By demonstrating the phenotypic and genotypic conservation of metabolic properties across three phylogenetic groups of Salmonella we have shed new light on the evolution of Salmonella serotypes .
|
Salmonella enterica is a remarkable bacterial species that causes Neglected Tropical Diseases globally . The burden of disease is greatest in some of the most poverty-afflicted regions of Africa , where salmonellosis frequently causes bloodstream infection with fatal consequences . The bacteria have the ability to colonise the gastrointestinal tract of a wide range of animals including reptiles . Direct or indirect contact between reptiles and humans can cause salmonellosis . In this study , we determined the prevalence and diversity of Salmonella in a collection of African venomous snakes for the first time . We showed that the majority of venomous snakes ( 91% ) in our study carry Salmonella , and used bacterial genomics to assign two thirds of isolates to the S . enterica subspecies enterica which is associated with human salmonellosis . We identified two evolutionary groups of S . enterica subspecies enterica that display distinct metabolic profiles with infection-relevant carbon sources . Our findings could have a broad significance because venomous snakes can move freely around human dwellings in tropical regions of the world such as Africa , and could potentially shed contaminated faecal matter onto surfaces and into water supplies .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"and",
"discussion"
] |
[
"taxonomy",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"microbiology",
"vertebrates",
"animals",
"salmonellosis",
"bacterial",
"diseases",
"reptile",
"genomics",
"phylogenetics",
"data",
"management",
"enterobacteriaceae",
"reptiles",
"pharmacology",
"bacteria",
"bacterial",
"pathogens",
"infectious",
"diseases",
"computer",
"and",
"information",
"sciences",
"antimicrobial",
"resistance",
"medical",
"microbiology",
"microbial",
"pathogens",
"salmonella",
"enterica",
"evolutionary",
"systematics",
"salmonella",
"animal",
"genomics",
"snakes",
"eukaryota",
"squamates",
"genetics",
"microbial",
"control",
"biology",
"and",
"life",
"sciences",
"genomics",
"evolutionary",
"biology",
"amniotes",
"organisms"
] |
2019
|
The diversity, evolution and ecology of Salmonella in venomous snakes
|
Malaria is a global health concern caused by infection with Plasmodium parasites . With rising insecticide and drug resistance , there is a critical need to develop novel control strategies , including strategies to block parasite sporogony in key mosquito vector species . MAPK signaling pathways regulated by extracellular signal-regulated kinases ( ERKs ) and the stress-activated protein kinases ( SAPKs ) c-Jun N-terminal kinases ( JNKs ) and p38 MAPKs are highly conserved across eukaryotes , including mosquito vectors of the human malaria parasite Plasmodium falciparum . Some of these pathways in mosquitoes have been investigated in detail , but the mechanisms of integration of parasite development and mosquito fitness by JNK signaling have not been elucidated . To this end , we engineered midgut-specific overexpression of MAPK phosphatase 4 ( MKP4 ) , which targets the SAPKs , and used two potent and specific JNK small molecule inhibitors ( SMIs ) to assess the effects of JNK signaling manipulations on Anopheles stephensi fecundity , lifespan , intermediary metabolism , and P . falciparum development . MKP4 overexpression and SMI treatment reduced the proportion of P . falciparum-infected mosquitoes and decreased oocyst loads relative to controls . SMI-treated mosquitoes exhibited no difference in lifespan compared to controls , whereas genetically manipulated mosquitoes exhibited extended longevity . Metabolomics analyses of SMI-treated mosquitoes revealed insights into putative resistance mechanisms and the physiology behind lifespan extension , suggesting for the first time that P . falciparum-induced JNK signaling reduces mosquito longevity and increases susceptibility to infection , in contrast to previously published reports , likely via a critical interplay between the invertebrate host and parasite for nutrients that play essential roles during sporogonic development .
The etiologic agents of malaria are protozoan parasites in the genus Plasmodium and are responsible for 216 million new cases and 445 , 000 deaths worldwide in 2016 [1] . Artemisinin-based combination therapies ( ACTs ) have been adopted as first-line treatments of uncomplicated and severe Plasmodium falciparum malaria in many countries with concomitant reductions in the global malaria burden [2] . Unfortunately , artemisinin-resistant malaria parasites have been detected in five countries in Southeast Asia [3] and spread of these strains to Africa or the Indian subcontinent could be disastrous . Anopheles stephensi , one of the major vectors of malaria in the Indian subcontinent and Middle East , is well adapted to urban areas and feeds aggressively on humans . The appearance of A . stephensi in Djibouti , Horn of Africa , has been linked to a recent resurgence of severe Plasmodium falciparum malaria [4] . More recently , A . stephensi has been detected in Sri Lanka , where it has never been reported , raising concerns regarding vulnerability of this country to reintroduction of malaria [5] . Therefore , continued efforts in the development of novel strategies and tools to curb malaria transmission , including those focused on the mosquito host , are still required . During sporogony , parasites encounter an array of impediments within the mosquito that can limit infection . Innate immune pathways include the Toll signaling cascade [6] , the Janus kinase/signal transducers and activators of transcription ( JAK/STAT ) [7] , and the immune deficiency ( IMD ) pathway [8] . Parasite activation of these signaling pathways results in the synthesis of antimicrobial peptides , reactive nitrogen and oxygen species ( RNOS ) , and other immune factors ( e . g . , TEP1 , APL1 , LRIM1 , LRRD7 ) that are anti-parasitic [9 , 10] . Signaling proteins and pathways that finely tune mosquito defense against parasite infection include the mitogen activated protein kinases ( MAPKs ) [11 , 12] , insulin/insulin-like growth factor signaling ( IIS ) [13–16] , and the transforming growth factor ( TGF ) -β signaling pathway [17] . These pathways regulate mosquito NO synthase ( NOS ) and production of RNOS [16 , 18 , 19] as well as intermediary metabolism and epithelial barrier function in the mosquito midgut to control parasite infection [12 , 16 , 20–22] . The stress-associated protein kinases ( SAPKs ) , including the c-Jun N-terminal kinases ( JNKs ) and p38 MAPKs , along with the extracellular signal-regulated kinases ( ERKs ) are members of the mitogen activated protein kinase ( MAPK ) superfamily . Mammalian JNK1 , JNK2 , and JNK3 are encoded by three different genes , all of which are associated with several alternatively spliced products ( e . g . , JNK1A ) . The single JNK ortholog in Drosophila melanogaster , Basket , triggers apoptosis in embryonic epithelia during normal development as well as in response to γ-irradiation [23] . Unlike D . melanogaster , the Anopheles gambiae genome encodes two JNK orthologs AGAP009461 and AGAP009460 that are homologous to mammalian JNK1 and JNK3 , respectively [24] . The A . stephensi genome encodes JNK1 ( ASTE007551 ) and JNK3 ( ASTE007552 ) that are in 1:1 orthology with A . gambiae JNK1 and JNK3 . JNK signaling has been implicated in mosquito defense against malaria parasites . In A . stephensi , human insulin-like growth factor 1 ( IGF1 ) induces NOS expression to mediate inhibition of P . falciparum development via enhanced JNK activation in the midgut epithelium [20] . In A . gambiae , JNK activation has been linked to the regulation of heme peroxidase 2 and NAPDH oxidase 5 , both of which function with NOS to opsonize parasites , leading to TEP1-mediated complement-like elimination in the mosquito host [25] . Other work has suggested that the P . falciparum protein Pfs47 , expressed in the ookinete stage , disrupts A . gambiae JNK signaling to facilitate parasite evasion of the immune response and enhance parasite survival in the mosquito [26] . In related work , RNAi-dependent gene silencing of the A . gambiae ortholog of puckered , a MAPK phosphatase ( MKP ) that negatively regulates Basket in Drosophila , enhanced resistance against Plasmodium berghei infection in A . gambiae [27] . Gene products that impact Plasmodium development make attractive targets for genetic modification to enhance immune resistance against malaria parasite infection . Indeed , overexpression of the IIS kinase Akt in the A . stephensi midgut epithelium resulted in complete refractoriness to P . falciparum infection in this mosquito host [13] . However , genetic manipulation can negatively impact key fitness traits such as lifespan and reproductive output . While this may be effective at reducing mosquito “infective lifespan” and preventing parasite development , modifications that decrease lifetime fecundity also limit the fitness of modified mosquitoes relative to unmodified natural populations . Perhaps more importantly , however , high fitness costs can reduce the efficiency of genetic drive mechanisms to fix transgenes into mosquito populations , preventing successful replacement of susceptible mosquito populations with those that are resistant to parasite infection . Evidence from D . melanogaster has demonstrated that moderate inhibition of the JNK signaling cascade , which regulates lifespan and immunity through midgut homeostasis , can mitigate the fitness costs of enhanced immunity . JNK activity induces stem cell proliferation in the fly midgut , but chronic signaling contributes to loss of tissue homeostasis in aged flies , particularly when exposed to high stress [28] . Reducing JNK activity prevented the age-associated changes in midgut physiology and improved resistance to stress . These observations indicated that age-related gut epithelium homeostasis , lifespan , and immune function are regulated , at least in part , by JNK signaling , physiology that is likely evident in related Diptera , including mosquito vector species in the genus Anopheles . To test this hypothesis , we utilized genetic manipulation of midgut SAPK signaling and potent JNK-specific small molecule inhibitors ( SMIs ) to test the hypothesis that parasite resistance and mosquito fitness are coordinately regulated by this pathway in A . stephensi . Both strategies inhibited JNK activity in the midgut epithelium , reduced the proportion of P . falciparum-infected mosquitoes and decreased oocyst loads relative to controls . SMI-treated mosquitoes exhibited no difference in lifespan compared to controls with modest effects on fecundity , whereas genetically manipulated mosquitoes exhibited increased longevity and reduced fecundity with increasing JNK signaling inhibition . Further , manipulation of JNK signaling with SMIs suggested that P . falciparum increases the accessibility of previously identified essential nutrients from its mosquito host via activation of this pathway . These results provide new insights into JNK regulation of mosquito physiology and vector competence and elucidate new mechanisms whereby mosquito life history traits are intimately connected with resistance to parasite infection .
JNK signaling increases with age in the midgut and is required for stress tolerance in midgut epithelial cells in D . melanogaster [28] . To determine if the same pattern was evident in mosquitoes , phosphorylated JNK ( pJNK ) levels were assessed in the head , thorax , abdomen , and midgut tissues of female A . stephensi at 7 , 14 , and 21 days-post-emergence . Relative pJNK1 and pJNK3 levels varied among mosquito tissues , but contrary to observations in D . melanogaster , changed little with age ( Fig 1 ) . To determine whether JNK was activated in response to malaria parasite infection , female A . stephensi were fed a P . falciparum-infected blood meal and midgut JNK phosphorylation levels were quantified at 30 min and 3 h post-blood feeding . The results revealed a trend towards increased JNK1 phosphorylation in mosquitoes fed P . falciparum-infected red blood cells ( RBCs ) at 30 min post-feeding compared to mosquitoes fed RBCs alone , and a significant increase in JNK1 phosphorylation at 3 h post-feeding ( Fig 2 ) . JNK3 phosphorylation levels remained at baseline for both time points ( Fig 2 ) . For MAPKs to be active , both the threonine and tyrosine residues in the conserved signature sequence threonine-x-tyrosine ( TxY ) within the activation motif must be phosphorylated , typically by an upstream MAPK kinase [11] . Thus , removal of a single phosphate from either motif is sufficient for MAPK inactivation . The MKPs , which are dual-specificity phosphatases ( DUSPs ) and members of the protein tyrosine phosphatase ( PTP ) superfamily , render MAPKs inactive by catalyzing the removal of both phosphates from the activation motif [29–31] . All MKPs contain a highly conserved C-terminal catalytic domain with the extended consensus signature motif DX26VLVHCX ( A/M ) G ( V/I ) SRSX5AYL with the residues in bold essential for catalysis [30 , 32] . Six MKPs have been identified in the D . melanogaster genome , all of which contain the extended conserved consensus motif [33–36] . The Anopheles MKPs also possess the highly conserved extended active site sequences and residues required for MKP catalysis , suggesting that the encoded proteins are enzymatically active ( Fig 3A ) . MKPs are grouped as class I , II , III or atypical MKPs based on subcellular localization and substrate specificities ( summarized in Table 1 ) [37] . Phylogenetic analysis revealed that A . gambiae AGAP012237 and A . stephensi ASTE011534 grouped with class II MKPs , and that AGAP004353 and ASTE002907 grouped with class III MKPs . AGAP002108 and ASTE004228 aligned closely with DUSP12 , and AGAP009903 and ASTE009886 aligned closely with DUSP19 , both of which are considered atypical DUSPs ( Fig 3B ) . The mosquito MKP genes have 1-to-1 orthology with D . melanogaster genes and , therefore , are named similarly ( Table 1 ) . To investigate whether mosquito MKPs regulate the activation of MAPKs , ASE cells were transiently transfected with V5-tagged plasmids encoding the complete protein sequence of A . gambiae MKP3 , MKP4 , or MKP5 , which are 75–85% identical in amino acid sequence to the A . stephensi orthologs ( not shown ) . Given that D . melanogaster MKP5 ( Puckered ) is both activatable and active in human embryonic kidney ( HEK293 ) cells [38] , highly conserved patterns of activity among heterologous Anopheles MKPs were expected among congeners . At 24 h post-transfection , cells were exposed to 10 μg E . coli lipopolysaccharide ( LPS ) for 15 min and cell lysates were assayed by western blotting to measure levels of phosphorylated ERK , JNK , and p38 MAPK [24] ( Fig 4 ) . In ASE cells , pJNK3 is detectable with anti-pJNK 1/2 antibody , whereas pJNK1 is faint or undetectable for reasons that are unclear to us , so the following observations are presented in this context . In ASE cells transfected with an empty vector , LPS treatment was associated with reduced ERK phosphorylation and increased p38 MAPK and JNK3 phosphorylation ( e . g . , fold changes of 0 . 69 , 12 . 85 and 7 . 66 , respectively , as indicated; Fig 4 ) . In the presence of MKP3 overexpression , LPS induced higher ERK phosphorylation and lower relative p38 and JNK3 phosphorylation levels than observed in the empty vector controls . With MKP4 overexpression , the effect of LPS on ERK phosphorylation was similar to control , but the relative fold changes in p38 and JNK3 phosphorylation in response to LPS were lower than those observed in the controls . With overexpression of MKP5 ( Puckered ) , relative p38 and JNK3 phosphorylation levels after LPS treatment were also reduced relative to controls , but MKP5 overexpression notably reduced ERK phosphorylation after treatment with LPS . Thus , out of the three phosphatases , MKP4 had the most desired characteristics , with minimal effect on ERK phosphorylation and reductions in inducible p38 MAPK and JNK3 phosphorylation compared to controls . Further , these data are consistent with observations that D . melanogaster MKP4 negatively regulates JNK activation , but not that of ERK [36] . Based on the effects of MKP4 overexpression in A . stephensi cells in vitro ( Fig 4 ) , we genetically engineered A . stephensi to overexpress MKP4 in the midgut epithelium ( CP-AsMKP4-HA ) to examine regulation of longevity , fecundity , immunity , and P . falciparum infection . A total of 10 transgenic A . stephensi lines overexpressing MKP4 under the control of the midgut-specific CP promoter ( Fig 5A and 5B ) were generated . Two lines ( M3 and M4 ) with the highest levels of transgene mRNA and protein expression were selected for further analyses . While both lines had robust transgene expression , the M3 line expressed significantly more MKP4 protein than did the M4 line ( S1 Fig ) . Transcript expression of CP-AsMKP4-HA was localized to the midgut in both lines and induced by blood feeding in adult female transgenic mosquitoes ( Fig 5C and 5D ) . Expression of the CP-AsMKP4-HA protein was similarly detected only in the midgut of both lines and not in the carcass ( Fig 5E ) . The M3 line expressed MKP4 protein in a manner typical for transgene expression driven by the carboxypeptidase promoter in A . stephensi . Specifically , MKP4 protein expression in the M3 line was increased for up to 24 h after the bloodmeal , after which it returned to basal levels ( Fig 5F ) . An increase in MKP4 protein was also observed at 72 h but this was highly variable among the replicates . In contrast , MKP4 protein expression in the M4 line did not increase following the bloodmeal , but instead remained consistent with the non-blood fed background level of MKP4 protein expression ( Fig 5F ) . It is important to note that expression levels between M3 and M4 cannot be directly compared because independent immunoblots were performed . However , as noted above the M3 line consistently expressed more MKP4 protein than the M4 line ( S1 Fig ) . These differences in protein expression between the M3 and M4 lines likely account for why , in the following assays , M3 phenotypes were significantly different from control and largely mirrored JNK inhibitor assays , whereas M4 phenotypes were typically not different from control . Finally , results from sequencing inverse PCR fragments revealed that the transgene did not insert into any known or predicted A . stephensi genes for either the M3 or M4 lines ( S2 Fig ) . Both lines were maintained as hemizygous by outcrossing female mosquitoes in each generation to non-transgenic male A . stephensi . The highly selective , potent inhibitor JNK-IN-8 forms a covalent bond with a conserved cysteine residue in JNK [39] , whereas TCS JNK 6o is an ATP-competitive pan-JNK inhibitor [40] . The residues forming the ATP binding site are largely conserved in A . stephensi JNK1 and JNK3 , but the conserved cysteine is missing in JNK3 ( S3 Fig ) . To test the activity of these small molecule inhibitors ( SMIs ) against JNK signaling in A . stephensi , 3-day old female A . stephensi were provided artificial blood meals supplemented with 1 μM JNK-IN-8 , 1 μM TCS JNK 6o , or an equivalent volume of diluent ( dimethyl sulfoxide , DMSO ) added to the blood meal as a control . At 1–3 h post feeding , 25 midguts from each group were dissected , pooled , and processed for western blotting for pJNK . Both SMIs significantly inhibited JNK phosphorylation ( Fig 6 ) , with differing effects of the two SMIs on JNK3 that might derive from the mechanisms described above . Specifically , midgut pJNK3 and pJNK1 levels in JNK-IN-8-treated mosquitoes were 82 . 8% and 72% of control levels , respectively , while pJNK3 and pJNK1 levels in TCS JNK 6o-treated mosquitoes were 59 . 6% and 49 . 8% of control levels , respectively . To determine whether MKP4 overexpression in the A . stephensi midgut could inhibit inducible JNK phosphorylation , we provided identical P . falciparum-infected blood meals to M3 and M4 transgenic lines and to non-transgenic controls and assessed pJNK levels in these insects relative to non-fed matched controls . This normalization was selected to allow us to determine whether MKP4 overexpression would be sufficient to alter pJNK signaling induced not only by infection , but also by ingestion of blood alone ( Fig 2 ) . In non-transgenic mosquitoes , JNK3 phosphorylation levels were significantly increased above non-fed control levels ( dotted line ) by 3 h post-feeding before returning to this baseline ( Fig 7 ) . In contrast , JNK3 phosphorylation declined significantly by 5 h post-feeding in M3 transgenic mosquitoes ( Fig 7 ) , consistent with MKP4 repression of inducible JNK3 phosphorylation in ASE cells ( Fig 4 ) , whereas JNK3 phosphorylation was not different from control levels at any time post-blood feeding in M4 transgenic mosquitoes . JNK1 phosphorylation levels were not different from control levels in non-transgenic or transgenic mosquitoes at any time post-blood feeding , although pJNK1 levels trended downward post-blood feeding relative to non-fed control levels only in the M3 transgenic line ( Fig 7 ) . Additionally , as with MKP4 overexpression in ASE cells , M3 transgenic mosquitoes overexpressing MKP4 in the midgut did not display increased p38 phosphorylation ( S4A Fig ) and ERK phosphorylation was unaffected ( S4B Fig ) following a P . falciparum-infected bloodmeal relative to non-transgenic controls . Based on previous observations from D . melanogaster , we hypothesized that exposure to JNK SMIs in weekly blood meals should extend mosquito lifespan relative to controls [28 , 41] . However , an analysis of median lifespans in 11 separate cohorts of A . stephensi provided weekly blood meals supplemented with 1 μM JNK-IN-8 or 1 μM TCS JNK 6o indicated that treated mosquitoes were not significantly different from matched controls that received blood meals supplemented with an equivalent volume of diluent ( one-way ANOVA , P = 0 . 80; Table 2 ) . Accordingly , we speculated that exposure to JNK SMIs once per week was not sufficient to impact lifespan , but that perhaps sustained JNK inhibition could impact longevity . To test this , female MKP4 transgenic mosquitoes were provided blood meals three times a week to induce consistent overexpression of the MKP4 transgene in the midgut epithelium . Here , blood feeding extended median lifespan in three of four cohorts of M3 mosquitoes by an average of 5 . 7 days ( Fig 8A and 8C ) , but significantly reduced median lifespan in one M4 cohort and enhanced median lifespan by 4 . 5 days in only two out of six M4 cohorts ( Fig 8B and 8C ) . An analysis of these replicates indicated that there was no difference in median lifespan between M4 line and non-transgenic females , but median lifespan of the M3 line was significantly longer than that of non-transgenic females ( one-way ANOVA , P = 0 . 05 ) . In separate studies , we examined the effects of JNK SMIs and MKP4 overexpression on reproductive output . The proportions of A . stephensi females that laid eggs during the first gonotrophic cycle after feeding on blood meals containing 1 μM JNK-IN-8 or 1 μM TCS JNK 6o were not significantly different from female mosquitoes that received a blood meal supplemented with an equivalent volume of diluent as a control ( Fig 9A ) . However , the average clutch sizes from groups that received either JNK inhibitor were lower than the control group ( Fig 9B ) . Specifically , control mosquitoes laid an average of 28 . 1 eggs per female compared to 21 . 9 and 24 . 8 eggs per female in mosquitoes fed JNK-IN-8 ( P = 0 . 026 ) or TCS JNK 6o ( P = 0 . 052 ) , respectively ( Fig 9B ) , with no difference between the treatment groups . Lifetime fecundity of MKP4 transgenic and non-transgenic A . stephensi was measured by tallying the total number of eggs produced during female reproductive lifetime . M3 mosquitoes produced significantly fewer eggs per female relative to non-transgenic controls ( Fig 10A ) , whereas M4 female output did not differ from controls ( Fig 10B ) . M3 transgenic mosquitoes produced fewer eggs relative to non-transgenic females over their reproductive lifespans ( Fig 10C ) . In contrast , the M4 transgenic line did not differ significantly from non-transgenic controls ( Fig 10D ) . Thus , even with the increased survivorship of M3 transgenic A . stephensi relative to non-transgenic controls ( Fig 8A and 8C ) , the M3 line produced fewer eggs suggesting a physiological trade-off between lifespan extension and reproduction . Three-day old adult female A . stephensi mosquitoes were provided P . falciparum gametocyte-enriched blood meals supplemented with 1 μM JNK-IN-8 or 1 μM TCS JNK 6o or an identical blood meal supplemented with an equivalent volume of diluent ( DMSO ) as a control . At day 10 post-blood feeding , midguts were dissected and stained and the numbers of oocysts were directly counted . In all three replicates , mosquitoes fed an infectious blood meal with JNK-IN-8 had significantly fewer oocysts per midgut compared to control ( Table 3 ) . Mosquitoes given an infectious blood meal supplemented with TCS JNK 6o had significantly fewer oocysts relative to control in one of three replicates with downward trends in the other two replicates ( Table 3 ) . Prevalences of infection ( i . e . , the number of mosquitoes that had at least one oocyst ) were consistently and significantly lower in both JNK-IN-8- and TCS JNK 6o-treated females relative to controls as determined by Chi-square analysis ( Exp 1 , χ2 = 20 . 78 , df = 2 , P = 0 . 000031; Exp 2 , χ2 = 11 . 53 , df = 2 , P = 0 . 0031; Exp 3 , χ2 = 24 . 85 , df = 2 , P < 0 . 00001 ) . While our data suggested that JNK SMIs alter parasite development in A . stephensi through effects on mosquito cell signaling , we also examined the possibility that these inhibitors could target parasite viability directly . The P . falciparum genome encodes two MAPKs , Pfmap-1 and Pfmap-2 , that do not cluster phylogenetically with typical MAPKs . In particular , Pfmap-1 clusters with human ERK8 and Pfmap-2 is an atypical MAPK with no homology to ERK1/2 , JNK , and p38 MAPK proteins [42] . Nevertheless , we tested the effects of JNK SMIs on parasite asexual growth in vitro , a reasonable approach in the absence of a complete sexual stage culture system . Synchronized asexual P . falciparum parasites were treated with 0 . 1–10 μM JNK-IN-8 or TCS JNK 6o in vitro and their growth was evaluated at 48 and 96 h post treatment . Only when parasites were grown in the presence of high concentrations ( 10 μM ) of JNK-IN-8 were negative effects on asexual growth observed ( S5 Fig ) . Importantly , treatment with 1 μM JNK-IN-8 or 1 μM TCS JNK 6o , concentrations used in infectious blood meals for mosquitoes ( Fig 9 ) , had no significant effects on parasite growth in vitro relative to control . Therefore , we inferred that reductions in infection prevalence and intensity due to JNK SMI treatment were due to the effects on mosquito cell signaling and physiology and not due to direct effects of these inhibitors on parasite viability . Malaria parasite infection was also assessed in the M3 and M4 MKP4 transgenic lines . In both lines , MKP4 overexpression in the midgut epithelium significantly reduced the percentage of mosquitoes infected with P . falciparum relative to non-transgenic A . stephensi ( Fig 11 ) . The percentage of mosquitoes with one or more oocysts decreased from an average of 46 . 6% in non-transgenic A . stephensi to 28 . 8% in M3 females and 28 . 9% in M4 females ( Fig 11A ) . Infection intensity , however , was only significantly reduced in M3 females relative to non-transgenic A . stephensi ( Fig 11B ) . The JNK signaling pathway is essential for the production of antimicrobial peptides [43 , 44] , the induction of NOS and production of NO [45 , 46] , and activation of the complement-like system [26] . Thus , we measured midgut expression levels of anti-parasite immune genes involved in these processes following blood meals containing P . falciparum gametocytes supplemented with 1 μM JNK-IN-8 or 1 μM TCS JNK 6o or an identical blood meal supplemented with an equivalent volume of diluent as a control . In control mosquitoes , transcript levels remained unchanged between 3 and 24 h ( Fig 12 ) . However , when mosquitoes were treated with 1 μM JNK-IN-8 , transcript levels for APL1 and LRIM1 were significantly reduced from 3 h to 24 h , while treatment with 1 μM TCS JNK 6o reduced transcript levels for APL1 , LRIM1 , and TEP1 from 3 h to 24 h after feeding ( Fig 12 ) . APL1 , LRIM1 , and TEP1 are transcriptionally controlled by Rel2 [9] and Rel1 [47] , NF-κB transcription factors that function downstream of Toll and Immune Deficiency ( IMD ) signaling , respectively . The IMD pathway is also networked with TAK1 , a MAP3K that is an upstream activator for JNK . The Rel2 orthologue in Drosophila , Relish , modulates the duration of JNK signaling and output in response to Gram-negative infections by inducing the proteasomal degradation of TAK1 [48] . Accordingly , we utilized NF-κB-dependent promoter-luciferase reporter constructs to investigate the extent of crosstalk between NF-κB and JNK signaling in the context of JNK inhibition in A . stephensi cells . ASE cells were pre-treated with 1 μM JNK-IN-8 , 1 μM TCS JNK 6o , or mock-treated for 1 h and then stimulated with LPS prior to assay of cecropin , defensin , and gambicin promoter-reporter activities . LPS induction of promoter activity was expected [12 , 49] and observed ( Fig 13 ) . However , inhibitor pre-treatment prior to LPS stimulation had no effect on NF-κB-dependent promoter activities ( Fig 13 ) , generally affirming effects of JNK SMIs on innate responses in vivo ( Fig 12 ) . We also quantified midgut anti-parasite gene expression in M3 and M4 MKP transgenic A . stephensi females and in non-transgenic controls fed identical P . falciparum-infected blood meals . Notably , there were no significant differences in transcript levels between 3 h and 24 h post-infection in either M3 or M4 mosquitoes ( Fig 14 ) . While MKP4 inhibited inducible phosphorylation of both JNK and p38 MAPK ( Figs 4 and S4 ) , our previous studies showed that specific inhibition of p38 MAPK increased midgut expression of LRIM1 , LRRD7 , NOS , TEP1 , and APL1 [12] , indicating that the effects of MKP4 overexpression are attributable to inhibition of JNK signaling , but not p38 MAPK signaling , in vivo in A . stephensi . Collectively , these observations and our in vivo ( Fig 12 ) and in vitro data for JNK SMIs ( Fig 13 ) suggested that transcriptional regulation of anti-parasite genes does not account for observed reductions in P . falciparum development in A . stephensi associated with JNK inhibition . Although JNK inhibition was not associated with enhanced NF-κB-mediated anti-parasite defenses in A . stephensi , studies in D . melanogaster suggested that other JNK-specific alterations to the midgut might be associated with enhanced resistance . Specifically , chronic JNK signaling in aged flies can lead to loss of tissue homeostasis but moderate inhibition of JNK reduced age-related midgut dysplasia and extended lifespan [50] . We previously reported that resistance to P . falciparum infection in A . stephensi can be associated with increased midgut epithelial integrity [14 , 16 , 49] , suggesting that JNK SMIs might enhance resistance by increasing midgut integrity . To test this , we modified a previous protocol to test midgut permeability with blood meals containing fluorescent particles [14] but without parasites to exclude confounding effects on the midgut epithelium related to parasite infection and invasion . Relative to controls , the midgut was significantly more permeable in mosquitoes fed blood meals supplemented with 1 μM JNK-IN-8 , whereas there was no effect of TCS JNK 6o ( Fig 15 ) , suggesting that JNK SMIs do not enhance resistance by consistently improving midgut barrier integrity . To examine this further , we tested midgut barrier integrity in MKP4 transgenic mosquitoes with the same protocol . Here , MKP4 overexpression did not alter midgut permeability in either transgenic line relative to control ( Fig 15 ) , again suggesting that JNK inhibition does not enhance resistance to P . falciparum infection by enhancing midgut barrier integrity . We previously used metabolomics to define the nature of midgut-associated insulin-like peptide ( ILP ) regulation of P . falciparum resistance in A . stephensi [22] , so this approach was chosen to help us define the nature of parasite resistance associated with JNK inhibition . While our data suggested that MKP4 activity is likely specific to JNK signaling in the A . stephensi midgut , we could not exclude the potential for MKP4 effects on p38 MAPK signaling ( Figs 4 and S4 ) , so we analyzed the effects of JNK SMIs on the metabolome of the A . stephensi midgut epithelium to more clearly define the effects of JNK inhibition . For these studies , female A . stephensi were provided blood meals supplemented with 1 μM JNK-IN-8 or 1 μM TCS JNK 6o or an identical meal supplemented with an equivalent volume of diluent as a control . We selected 24 h post-feeding based on effects on midgut gene expression ( Fig 12 ) to dissect and prepare midgut samples for analysis . Provision of JNK SMIs increased lactate levels ( JNK-IN-8 ) and pyruvate levels ( TCS JNK 6o ) in the A . stephensi midgut relative to control ( S6 Fig ) , whereas our previous studies showed that inhibition of p38 MAPK with the SMI BIRB796 increased lactate and decreased pyruvate [12] , separating the effects of JNK and p38 MAPK on A . stephensi midgut intermediary metabolism . With GC-MS/MS of midguts from control mosquitoes and those treated with JNK SMIs , we detected and identified 135 metabolites ( Fig 16 ) . A large majority of compounds showed differential abundance in JNK SMI-treated versus control mosquitoes ( 85% for JNK-IN-8 and 81 . 5% for TCS JNK 6o; Fig 16A ) . Similar numbers of metabolites showed higher and lower abundances relative to controls for both JNK SMIs ( 60% and 66 . 4% higher in red , 40% and 33 . 6% lower in green for JNK-IN-8 and TCS JNK 6o , respectively , Fig 16A ) . Linear regression of the relative concentrations of metabolites for both inhibitors ( Fig 16B ) indicated that treatment with these two JNK SMIs resulted in analogous changes in metabolites and perhaps analogous pathways . Metabolites with different abundances in JNK SMI-treated mosquitoes versus controls ( Fig 17A ) along with their relative concentrations were analyzed using a Metabolite Set Enrichment Analysis ( MSEA ) . Relative concentrations of metabolites were entered for quantitative enrichment analysis ( QEA ) using the metabolic pathway-associated metabolite set library and the GlobalTest package [51] . The QEA was performed using a generalized linear model to estimate a Q-statistic for each metabolite set , which described the correlation between compound concentration and treatment . The results are summarized as the average Q statistics for each metabolite in the input set ( Fig 17B ) . Pathways significantly enriched were the pantothenate and coenzyme A ( CoA ) pathways , sucrose and nucleotide sugar metabolism , steroid synthesis , pentose phosphate shunt , fatty acid metabolism , pyruvate and galactose metabolism , and glycolysis and pyruvate metabolism ( Fig 17B ) . The significant decrease in the ratio of active-to-total midgut pyruvate dehydrogenase complex ( PDHC ) suggested that JNK SMIs were inhibiting PDHC activity ( S6 Fig ) . This was consistent with over-representation of the pantothenate pathway via metabolite-dependent inhibition of PDHC derived from increased acetyl CoA from fatty acid beta-oxidation to generate ketone bodies ( vide infra ) . In parallel , we performed Pathway Analysis with normalized compound names using KEGG and PubChem databases for compound identification and the D . melanogaster pathway library . A detailed list of the pathways identified and their relative impacts are shown in Table 4 and Fig 18 , respectively . The most over-represented significant pathways ( in decreasing order ) were steroid biosynthesis , fatty acid metabolism , pentose phosphate shunt , inositol phosphate metabolism and ketogenesis . Thus , both analyses confirmed and extended the same conclusions: treatment with JNK SMIs was associated with over-representation of cholesterol , stigmasterol and ergosterol synthesis and over-representation of the CoA pathway , increased glucose flux through the pentose phosphate pathway ( to sustain not only antioxidant defenses but also nucleic acid synthesis ) and glucuronidation reactions at the expense of decreasing the flux through glycolysis . The abundances of all TCA cycle intermediates , including alpha-ketoglutarate , succinic acid , and citric acid were lower than control levels , suggesting reduced activity of the TCA cycle . Excess acetyl CoA ( from glucose and ketogenic amino acids ) not utilized in either the TCA cycle or cholesterol/sterol synthesis would be shunted to the formation of ketone bodies . A significantly lower concentration of most amino acids ( 16 of 17 detected ) was observed likely because of proteolysis followed by the occurrence of anaplerotic reactions given the lower glucose flux to the TCA . In this regard , treatment with JNK SMIs is associated with catabolism of substrates other than glucose , including proline and trehalose , critical fuels for intermediary metabolism in mosquitoes [52] . The enrichment of pantothenate and CoA biosynthesis pathways suggested a mechanism for lifespan extension by inhibition of JNK signaling . Pantothenate kinase is the rate limiting enzyme for the conversion of pantothenate to CoA and previous work implicated upregulated pantothenate kinase-1 ( pnk-1 ) as a critical mediator of lifespan extension in long-lived C . elegans daf-2 ( insulin receptor ortholog ) mutants [53 , 54] . Further , knockdown of pnk-1 using RNAi led to a threefold increase in the aging rate of C . elegans and dramatically shortened adult lifespan [53 , 54] . Accordingly , we hypothesized that an increase in midgut pantothenate kinase would activate FOXO and inhibit IIS , which we have previously associated with lifespan extension and P . falciparum resistance in PTEN-overexpressing A . stephensi [14] . To assess this possibility in the context of a signaling manipulation that extended A . stephensi lifespan ( Fig 8 ) , we examined midgut mRNA expression of A . stephensi pantothenate kinase ( PANK ) and PTEN in the M3 and M4 transgenic lines at 2–72 h post blood meal ( Fig 19 ) . We observed significant increases in PANK at 2 h and 24 h post-blood meal and PTEN at 36 h post blood meal in the M3 line ( Fig 19A and 19B ) , an effect that was consistent with increased MKP4 expression in this line ( Figs 5 and S1 ) . In contrast , neither PANK nor PTEN were induced in the midgut of the M4 line ( Fig 19C and 19D ) consistent with reduced resistance to P . falciparum ( Fig 11 ) and lack of consistent lifespan extension ( Fig 8 ) observed in this line relative to the M3 line . In sum , these data provided a novel link between JNK inhibition and IIS in the regulation of these life history traits .
With complementary experimental approaches , we have demonstrated that moderate inhibition of JNK signaling in the A . stephensi midgut extends lifespan and enhances resistance to the human malaria parasite P . falciparum . Resistance was independent of effects on NF-κB-dependent innate immunity , adding to the list of similar observations in A . stephensi that highlight the importance of alternative signaling pathways and intermediary metabolism in mediating anti-parasite resistance [14 , 20 , 22] . The phenotypic effects of JNK inhibition may be due in part to inhibition of IIS in the A . stephensi midgut , which we have previously associated with lifespan extension and resistance to parasite infection and which also reaffirms our observations of IIS-MAPK signaling cross-talk in the A . stephensi midgut [20 , 22] . Importantly , our data contrast with previous studies in A . gambiae reporting that RNAi-mediated silencing of JNK and associated genes enhanced susceptibility , rather than resistance , to Plasmodium berghei infection [27] . Here , the authors assumed but did not confirm that hemipterous ( MAP2K7 ) activates JNK signaling , they examined only a single A . gambiae JNK isoform ( JNK1 ) despite prior identification of two JNK isoforms in this mosquito species [24] , and assumed but did not confirm that Puckered specifically dephosphorylates JNK1 in A . gambiae . In A . stephensi cells , MKP5 ( Puckered ) decreased inducible activation of ERK , JNK , and p38 MAPK signaling ( Fig 4 ) , indicating that confirmation of MKP target specificity is critical for subsequent bioassays . In a follow on study , the same laboratory added JNK-interacting protein ( JIP ) , with a single ortholog ( JIP1 ) in A . gambiae , to their assays . Here , the authors reported that silencing of JIP1 , JNK ( JNK1 ) , Fos , Jun , and Puckered had no effect on the prevalence or intensity of infection of A . gambiae with P . falciparum NF54 , again without confirmation of protein-protein signaling interactions [26] . Notably , our colleagues asserted that P . falciparum Pfs47 was responsible for suppression of JNK signaling in A . gambiae , although this conclusion was based solely on patterns of infection with wild type and Pfs47 knockout parasites in lacZ and JNK signaling RNAi-silenced mosquitoes [26] . No confirmatory analyses of JNK signaling or altered protein phosphorylation in any tissue were presented in support of these studies , making interpretations of these findings , as with earlier studies , difficult . Subsequent inferences from analyses of laboratory P . falciparum strains that specific Pfs47 mutations mediate “lock-and-key” relationships with geographically sympatric Anopheles spp . hosts were not supported by analyses of Pfs47 sequences in field isolates of P . falciparum and patterns of infections of sympatric mosquitoes with these isolates [55] . From our metabolomic analyses of JNK SMI-treated A . stephensi , pathways that were most enriched in the midgut included the pantothenate and CoA synthesis pathways and fatty acid metabolism . Accompanying these enrichments , a reduced flux through the TCA cycle was evidenced by lower concentrations of all intermediates relative to controls . These changes are supported by increased lactate and alanine per glucose , higher levels of pantothenic acid , increased fatty acid beta-oxidation , and lower levels of TCA cycle intermediates , which are maintained by anaplerotic reactions ( see levels of key amino acids ) . A similar “low glycolytic state” versus “increased fatty acid oxidation state” was reported by us as functionally associated with reduced P . falciparum infection in A . stephensi [22] . Furthermore , transcription of A . stephensi PANK , which regulates the first critical step in CoA synthesis , was significantly upregulated in the midgut epithelium of the M3 line of MKP4 transgenic mosquitoes . PANK gene expression is directly controlled by the IIS transcription factor FOXO [53 , 54] . Active IIS phosphorylates FOXO , excluding it from the nucleus and preventing it from transcribing target genes such as PANK . Conversely , suppression of the IIS cascade , for example through enhanced PTEN activity , leads to increased FOXO-dependent transcriptional activity and , as demonstrated in numerous organisms [56 , 57] , including A . stephensi [14] , an increase in lifespan . While these metabolic shifts supported enhanced resistance and lifespan , they were also associated with significantly reduced fecundity , patterns consistent with PTEN-mediated extension of lifespan and enhanced resistance with reduced lifetime fecundity in A . stephensi [14] . However , JNK inhibition was not associated with altered or enhanced midgut barrier integrity–as observed in PTEN overexpressing mosquitoes [14]–that could be consistently associated with enhanced resistance to P . falciparum . Accordingly , inhibited IIS could not fully explain the phenotypes associated with repression of JNK signaling in A . stephensi . In this context , we reasoned that metabolic shifts in A . stephensi resulting from JNK inhibition might directly affect P . falciparum development in the mosquito host . Intriguingly , several recent studies strongly support the likelihood of such direct effects on P . falciparum development . William Trager , credited as the first individual to successfully continuously culture P . falciparum , also determined that panthothenate , as the precursor for CoA biosynthesis , was required for blood stage growth of several malaria parasite species , including P . falciparum [58] . This requirement is met by uptake of exogenous pantothenate because P . falciparum cannot synthesize pantothenic acid de novo [59 , 60] . Using P . yoelii as a model , Hart et al . [61] characterized a parasite pantothenate transporter PyPAT that , upon knockout , had no effect on asexual development and gametocyte formation , but that blocked ookinete development and the formation of oocysts and sporozoites in A . stephensi . More recently , Hart et al . characterized two putative pantothenate kinase genes , PyPanK1 and PyPanK2 , from P . yoelii [62] . Knockouts of either gene target completed normal asexual development and gametocyte formation in mouse erythrocytes , affirming previous observations that some malaria parasite species could utilize alternative host precursors for CoA synthesis . However , knockout parasites were severely deficient in ookinete development and completely unable to produce sporozoites in A . stephensi , indicating that both PanK1 and PanK2 are required for the development of ookinetes and oocysts . Hart et al . also characterized CoA pathway enzymes in P . yoelii , noting that de novo biosynthesis is essential for both blood and mosquito stages of parasite development [63] . Taken together , these studies demonstrate that malaria parasites must take up pantothenate from the host to complete their CoA synthesis and these processes are essential for sporogony in the mosquito host . In this light , the upregulation of panthothenate , PANK expression and CoA synthesis in the A . stephensi midgut epithelium in response to JNK inhibition would likely compete directly with extracellular sporogonic parasites for these essential resources , suggesting for the first time that shifts in intermediary metabolism of mosquito host tissues may directly and adversely affect essential parasite metabolism . With the observations that pantothenate and CoA synthesis are essential for malaria parasite growth and development in both the mammalian and mosquito hosts , there is compelling interest in the development of novel pantothenate analogs and CoA synthesis-targeted inhibitors [64] as antimalarial compounds . To this we can add that targeted , genetic manipulations of intermediary metabolism of the mosquito host to synergize with these antimalarial compounds could provide enhanced confidence that transmission blocking could be complete .
Drosophila MKP protein sequences were used to query the annotated A . gambiae genome [65] to uncover orthologues using Basic Local Alignment Search Tool ( BLASTp ) . As an additional search strategy , vertebrate MKP protein sequences were used to query the A . gambiae genome . From both searches , six A . gambiae genes encoding proteins with significant similarities to the Drosophila and vertebrate MKPs were identified and used to identify A . stephensi orthologues ( http://vectorbase . org ) . The extended consensus signature motifs of the MKP catalytic domain sequences from A . gambiae , A . stephensi , D . melanogaster , Homo sapiens , and Mus musculus were aligned using the MUltiple Sequence Comparison by Log-Expectation ( MUSCLE ) method . A neighbor-joining tree was constructed from the alignment using MEGA 5 . 05 software with the JTT model and 2000 bootstrap replications . JNK-IN-8 [39] and TCS JNK 6o [40] were obtained from Selleck Chemicals ( Houston , TX ) and Tocris Bioscience ( Minneapolis , MN ) , respectively . Antibodies used in this study were anti-GAPDH ( G9545; Sigma-Aldrich ) ( 1:10 , 000 ) , and anti-pJNK1/2 Thr183+Tyr185 ( 44682G; Thermo Fisher Scientific , Waltham , MA ) ( 1:3 , 000 ) . Anti-mouse IgG-peroxidase ( A9044; Sigma-Aldrich , St . Louis , MO ) ( GAPDH 1:20 , 000 and pJNK1/2 1:10 , 000 ) was used as a secondary antibody for immunoblotting . The immortalized , A . stephensi embryo-derived ( ASE ) cell line ( a gift from H . -M . Müller , EMBL ) was maintained in modified minimal essential medium ( MEM; Gibco , Invitrogen , Carlsbad , CA ) with 5% heat-inactivated fetal bovine serum at 28°C under 5% CO2 . A . stephensi mosquitoes were reared and maintained at 27°C and 80% humidity under a 12-hour light/dark cycle . Adult mosquitoes were maintained on 10% sucrose solution-soaked cotton pads and mice were used as a blood source for colony maintenance . Eggs were placed in water and larvae were fed a mixture of liquid food containing 2% w/v powdered fish food ( Sera Micron ) and baker’s yeast in a 2:1 ratio , and Game Fish Chow pellet food ( Purina ) . All mosquito rearing protocols were approved and in accord with regulatory guidelines and standards set by the Institutional Animal Care and Use Committee of the University of California , Davis . For in vivo studies , 3 day old female mosquitoes were kept on water for 24 h and then allowed to feed for 30 min on an artificial blood meal of uninfected human erythryocytes resuspended in phosphate-buffered saline ( PBS ) provided through a Hemotek Insect Feeding System ( Discovery Workshops , Accrington , UK ) . Chemical treatments were added to the erythrocyte-PBS mixture immediately prior to feeding ( typically 1–5μl per ml of erythrocytes-PBS ) with identical meals supplemented with an equivalent volume of diluent added for controls . To harvest cellular proteins from A . stephensi ASE cells , cells were pelleted , washed with cold PBS , pelleted again , and then lysed in cell lysis buffer ( 10 mM Tris-HCl pH 7 . 4 , 1mM EDTA , 100mM NaCl , 1mM NaF , 1mM EGTA , 2mM Na3VO4 , 20 mM Na4P2O7 , 0 . 1% SDS , 1% Triton X-100 , 0 . 5% sodium deoxycholate , 1 mM phenylmethylsulfonyl fluoride , 10% glycerol , 60 mg/mL aprotinin , 10 mg/ml leupeptin , 1 mg/ml pepstatin , 1 mg/ml calyculin A ) . Cellular debris was removed by centrifugation at 14 , 000 × g for 20 min at 4°C . The resulting supernatants were mixed with Laemmli sample buffer ( 125 mM Tris-HCl pH 6 . 8 , 10% glycerol , 10% SDS , 0 . 006% bromophenol blue , 130 mM dithiothreitol ) and the proteins were denatured at 95°C for 3 min prior to electrophoresis . Cellular proteins from mosquito midguts were processed as described previously [11] . Briefly , mosquito midguts were dissected into a protease inhibitor cocktail ( Sigma ) in ice-cold PBS and mixed to release blood , if any , by pipetting up and down . The midguts were washed in a filter column fitted with a fine mesh with the same PBS mixture until all the blood was removed . A final volume of the PBS mixture was added to loosen the midgut tissue from the filter and the tissue was transferred to a fresh tube and prepared for electrophoresis as described for cell culture lysates above . Proteins were separated on 10% SDS-PAGE polyacrylamide gels at 135 V for 1 h , 40 min and then transferred to nitrocellulose membranes ( BioRad Laboratories ) for 1 h , 15 min at 7 V . Membranes were blocked in nonfat dry milk ( 5% w/v ) in 1X Tris-buffered saline ( TBS; pH 7 . 0 ) containing 0 . 1% Tween ( TBS-T ) for 1 h at room temperature , and then reacted overnight in primary antibody at 4°C . The membrane was washed 3 times , 5 min each with 1X TBS-T followed by incubation with appropriate secondary antibody at 4°C overnight . The membrane was washed again three times , 5 min each with 1X TBS-T and then incubated in SuperSignal West Dura Extended Duration Substrate ( Pierce ) . The MultiDoc-It Imaging System and VisionWorks LS Image Acquisition software ( UVP ) were used to acquire membrane images . Densitometry analyses of antibody-bound proteins were performed using ImageJ software [66 , 67] . Levels of phosphorylated JNK1 and JNK3 in each treatment were normalized to GAPDH levels to control for protein loading . For in vitro assays , pJNK1 and pJNK3 levels in treated cells were normalized to the same in diluent-treated control cells , whereas midgut pJNK1 and pJNK3 levels in treated or transgenic mosquitoes were normalized to GAPDH levels in the same samples , then to levels in control or non-transgenic mosquitoes to derive fold changes in phosphorylation . Expression plasmids containing the complete coding sequences of A . gambiae MKP3 [Genbank: XM_320303] , MKP4 [XM_320933] , and MKP5 [XM_001688405] followed by a V5 tag in the pDream2 . 1/MCS vector ( Genscript , Piscataway , NJ ) were transfected into ASE cells using Effectene reagent ( Qiagen , Germantown , MD ) according to the manufacturer’s protocol . In brief , 1×106 cells in 1 . 6 ml medium were plated in 6-well tissue culture plates overnight at 28°C . At 24 h after plating , cells were transfected with 1 μg of plasmid DNA and incubated at 28°C . Cecropin , Defensin , and Gambicin promoter-reporter plasmids were transfected into ASE cells as described [68 , 69] . Briefly , 0 . 5×106 cells in 0 . 25 ml medium were plated in 48-well tissue culture plates overnight at 28°C . At 24 h post-transfection , cells were treated with 1 μM JNK-IN-8 or TCS JNK 6o or an equivalent volume of diluent as a control for 1 h , and then challenged with 100 μg/ml of LPS ( Escherichia coli serotype O26:B6; Sigma-Aldrich , St . Louis , MO ) . Luciferase activity was measured 24 h after immune challenge with the Dual-Glo system ( Promega , Madison , WI ) . Transgenic A . stephensi lines overexpressing MKP4 under the control of the midgut-specific carboxypeptidase ( CP ) promoter were created with an HA epitope at the carboxyl terminus ( AsteMKP4-HA ) to facilitate protein identification . The construct was inserted into the pBac[3XP3-EGFPafm] plasmid vector using AscI . The construct was sent to the University of Maryland Biotechnology Institute , Insect Transformation Facility ( UMBI-ITF ) for transformation of A . stephensi . A total of 10 EGFP-positive A . stephensi lines were generated . We chose two representative lines , M3 and M4 , based on strong ( M3 ) and moderate ( M4 ) midgut-specific transcript and protein expression patterns . Transgene insertion sites in the mosquito genome were identified by inverse PCR following the protocol of Buchholz et al . [70] . Transgenic mosquitoes were maintained as hemizygous lines by outcrossing the mosquitoes in each generation to colony non-transgenic A . stephensi in order to maximize genetic diversity . For experiments using JNK SMIs , female A . stephensi were maintained in cartons and provided artificial blood meals once per week for the duration of their lifespans . Each replicate was initiated with 300 mosquitoes and blood meals were supplemented with JNK-IN-8 , TCS JNK 6o , or an equivalent volume of diluent added to the blood meal as a control . Dead mosquitoes were counted three times per week and oviposition cups provided after each blood feeding . These experiments were replicated with separate cohorts of mosquitoes . Mosquitoes hemizygous for the CP-AsteMKP4-HA construct were mated with non-transgenic A . stephensi to generate 50% transgenic and 50% non-transgenic sibling mosquitoes . The resulting larvae were reared together under identical conditions and separated based on EGFP eye fluorescence of pupae under a fluorescent stereomicroscope . Mosquitoes were provided uninfected blood meals three times per week for the duration of adult life in addition to 10% sucrose ad libitum . Daily mortality for each treatment was recorded and dead mosquitoes were removed until all mosquitoes had perished . These experiments were replicated a minimum of three times using separate cohorts . Because of cost , the effects of JNK SMIs on fecundity were examined only for the first gonotrophic cycle . To this end , 3–5 day-old female A . stephensi were provided an artificial blood meal supplemented with JNK-IN-8 , TCS JNK 6o , or an equivalent volume of diluent added to the blood meal as a control . Mosquitoes that did not feed or only partially fed were discarded . At 48 h after feeding , females were individually housed in modified 50 ml conical tubes and provided water for oviposition . At 72–96 h after blood feeding , eggs were collected from each conical tube , washed onto filter paper and photographed . Eggs were then counted using ImageJ software [66 , 67] . To study the effects of transgene expression on lifetime fecundity , hemizygous CP-AsteMKP4-HA and non-transgenic controls were maintained under identical conditions . Briefly , groups of 150 transgenic and non-transgenic A . stephensi females were mated with 50 non-transgenic males . Each cage of 150 female mosquitoes was blood fed daily throughout their lifespans to ensure equal blood feeding success . Damp filter paper for oviposition was provided at 48 h after the first blood meal and replaced every 24 h if eggs were present . Eggs from transgenic and non-transgenic A . stephensi cages were counted using ImageJ software . This experiment was replicated a minimum of three times with separate cohorts . For mosquito infection studies , the NF54 strain of P . falciparum was initiated at 1% parasitemia in 10% heat-inactivated human serum and 6% washed red blood cells in RPMI 1640 with HEPES ( Gibco , Invitrogen ) and hypoxanthine . At days 15–17 , stage V gametocytes were evident and exflagellation was evaluated at 200× magnification with phase-contrast or modified brightfield microscopy the day before and day of feeding by observation of blood smears and before addition of fresh media . Female A . stephensi were provided water only ( no sucrose ) for 24 h and then fed a meal of P . falciparum-infected erythrocytes . Mosquitoes ( n = 150 per group ) were given 30 min to feed , after which non-fed and partially fed mosquitoes were removed , and those that were fully engorged were maintained on cotton pads soaked in 10% sucrose until day 10 post-infection . On day 10 , midguts were dissected and stained with 0 . 5% mercurochrome for visualization of P . falciparum oocysts by microscopy . These experiments were replicated using four separate P . falciparum cultures and separate cohorts of A . stephensi . For P . falciparum growth assays , NF54 cultures were synchronized with sorbitol and aliquots were plated in complete RPMI 1640 medium with HEPES , hypoxanthine , and 10% heat inactivated human serum . Parasites were treated with 0 . 1 , 1 , 10 μM JNK-IN-8 or TCS JNK 6o or an equivalent volume of diluent as a control and incubated at 37°C for 48 or 96 h before culture media were replaced with 10% formalin in RPMI 1640 . RBCs were stained with 10 μg/mL propidium iodide ( Sigma-Aldrich ) at room temperature for 1 h and infected cells counted with FACSCalibur flow cytometer ( BD Biosciences , San Jose , CA ) . These assays were replicated four times with separate parasite culture passages . Transgene expression patterns were examined in the midgut epithelium and carcass ( body minus midgut ) in non-transgenic and transgenic female A . stephensi ( 3–5 day old ) before and 24 h post blood feeding . Additionally , transgene transcripts were amplified in midgut tissue at 2 , 6 , 12 , 24 , 36 , 48 , and 72 h post-blood feeding in transgenic mosquitoes . In both experiments , total RNAs were isolated from the midguts and carcasses ( RNeasy; Qiagen ) and converted into cDNA ( High-Capacity cDNA Reverse Transcription Kit; Thermo ) per the manufacturer’s instructions Primers specific to MKP4 ( MKP4-HA-For: TCCGAAGTGTACCGTGAAGA; MKP4-HA-Rev: AGCGTAATCTGGCACATCGT ) and actin ( Actin-For: AGCGTGGTATCCTGACGCTGAAAT; Actin-Rev AACCTTCGTAGATCGGCACGGTAT ) were used . To investigate anti-parasite gene mRNA expression patterns , transgenic and non-transgenic female A . stephensi ( 3–5 day old ) were fed identical P . falciparum-infected blood meals that were untreated or treated with JNK SMIs or diluent and , at 3 h and 24 h post-blood feeding , 15 midguts were pooled and dissected into cold PBS . Total RNA was isolated and reverse transcribed into cDNA as above . Primers for A . stephensi APL1 , Defensin , LRIM1 , LRRD7 , NOS , and TEP1 were as previously described [14] . To investigate IIS regulation by MKP4 overexpression , we examined the mRNA expression of PANK and PTEN at the same time points described above for transgene expression using primer sets for PANK ( PANK-For: GCCAACCGTACCCGTTTAT; PANK-Rev: CGGAGATGCGCTTGTAGTT ) and PTEN ( PTEN-For: GCTTCACCAGTAATCGCAGTA; PTEN-Rev: GGTGGCCTAGCGTCTAAATTAT ) . Anti-parasite gene mRNA expression assays were performed using Maxima SYBR Green/ROX qPCR Master Mix ( Fermentas , Waltham , MA ) and an ABI Prism 7300 Sequence Detection System ( Applied Biosystems , Foster City , CA ) . The amplification conditions were as follows: initial denaturation at 95°C for 2 min , followed by 40 cycles of denaturation for 15 s at 95°C and annealing/extension at 60°C for 30 sec . Amplification results were analyzed by normalizing the data to the mosquito housekeeping gene ribosomal protein S7 . For PANK and PTEN qPCR expression analysis , Maxima SYBR Green/ROX qPCR Master Mix ( Fermentas , Waltham , MA ) was used with an Eppendorf RealPlex2 Mastercycler ( Eppendorf , Hauppauge , NY ) . The amplification conditions were as follows: initial denaturation at 95°C for 3 min , followed by 40 cycles of denaturation for 30 s at 95°C and annealing/extension at 60°C for 30 sec . Amplification results were analyzed by normalizing the data to the mosquito housekeeping gene ribosomal protein S7 . Three-day old female A . stephensi were maintained on water for 48 h and then allowed to feed for 30 min on artificial blood meals ( human erythrocytes in PBS ) with 2×106 magnetic fluorescent particles/ml ( 2 . 0–2 . 4 μm , Fluorescent Yellow Carboxyl Magnetic Particles FCM-2052-2; Spherotech , Lake Forest , IL ) . Blood meals with fluorescent particles were supplemented with equivalent volumes of 1 μM JNK-IN-8 , 1 μM TCS JNK 6o , or diluent ( DMSO ) as a control for SMI-treated mosquitoes . Non-fed mosquitoes were immediately removed after feeding . At 72 h post-feeding , samples of five pooled whole mosquitoes or five pooled dissected midguts were placed in sterile water , pulse-sonicated , and passed through a 35 μm filter to remove tissue debris . Samples were then transferred into 1 . 5 ml microcentrifuge tubes . Fluorescent particles were collected using the MagnaRack Magnetic Separation Rack ( Thermo Fisher Scientific ) and washed twice with sterile water , then collected into a final volume of 5–10 μl sterile water . AbsorbMax Black Sealing Film ( Sigma ) was cut to size and applied to the bottom of a diagnostic slide with white Teflon printed circles . The magnetic fluorescent particles were pipetted onto the circles and allowed to dry . Images were acquired with a stereoscope with fluorescence ( Olympus ) and processed and analyzed using Sketchbook Pro ( Autodesk , San Rafael , CA ) and ImageJ software . To estimate the number of particles that passed through the midgut epithelial tissue and into the body cavity , the numbers of particles in the five midguts were subtracted from each analyzed sample of five whole mosquitoes , removing the contribution of beads remaining in the midgut of whole body samples . The experiment was replicated with three separate cohorts of mosquitoes . Mosquitoes were provided artificial blood meals supplemented with 1 μM JNK-IN-8 , 1 μM TCS JNK 6o , or with an equivalent volume of diluent as a control . At 24 h after feeding , midguts were dissected from 100 mosquitoes in each treatment and control , collected into HEPES buffer with protease inhibitor cocktail in ice-cold PBS , and pulse sonicated . Samples were extracted following published protocols [71] . Thirty μl aliquots were extracted by 1 ml of degassed acetonitrile:isopropanol:water ( 3:3:2 , v/v/v ) at -20°C , centrifuged , decanted , and evaporated to complete dryness prior to a clean-up step with acetonitrile in water ( 1:1 ) . The cleaned extract was aliquoted into two equal portions and the supernatant was dried down again . Internal standards ( C08-C30 fatty acid methyl esters ) were added and the sample was derivatized by methoxyamine hydrochloride in pyridine and subsequently by N-methyl-N-trimethylsilyltrifluoroacetamide for trimethylsilylation of acidic protons . Data were acquired using the following chromatographic parameters [72]: Column: Restek corporation rtx5Sil-MS ( 30 m length x 0 . 25 mm internal diameter with 0 . 25 μm film of 95% dimethyl- and 5% diphenyl-polysiloxane ) ; Mobile phase: helium; Column temperature: 50–330°C; Flow-rate: 1 ml min-1; Injection volume: 0 . 5 μl; Injection: 25 splitless time into a multi-baffled glass liner; Injection temperature: 50°C ramped to 250°C by 12°C s-1; Gradient: 50°C for 1 min , then ramped at 20°C min-1 to 330°C , held constant for 5 min . Mass spectrometry parameters were as follows: a Leco Pegasus IV mass spectrometer was used with unit mass resolution at 17 spectra s-1 from 80–500 Da at -70 eV ionization energy and 1800 V detector voltage with a 230°C transfer line and a 250°C ion source . Data processing: Raw data files were preprocessed directly after data acquisition and stored as ChromaTOF-specific * . peg files , as generic * . txt result files and additionally as generic ANDI MS * . cdf files . ChromaTOF vs . 2 . 32 was used for data preprocessing without smoothing , 3 sec peak width , baseline subtraction just above the noise level , and automatic mass spectral deconvolution and peak detection at signal/noise levels of 5:1 throughout the chromatogram . Apex masses were reported for use in the BinBase algorithm . Resulting * . txt files were exported to a data server with absolute spectra intensities and further processed by a filtering algorithm implemented in the metabolomics BinBase database . The BinBase algorithm used the settings: validity of chromatogram ( <10 peaks with intensity greater than 10^7 counts per sec ) , unbiased retention index marker detection ( MS similarity >800 , validity of intensity range for high m/z marker ions ) , retention index calculation by 5th order polynomial regression . Spectra were cut to 5% base peak abundance and matched to database entries from most to least abundant spectra using the following matching filters: retention index window ±2 , 000 units ( equivalent to about ±2 sec retention time ) , validation of unique ions and apex masses ( unique ion must be included in apexing masses and present at >3% of base peak abundance ) , mass spectrum similarity fit to criteria dependent on peak purity and signal/noise ratios and a final isomer filter . Failed spectra were automatically entered as new database entries if s/n >25 , purity <1 . 0 and presence in the biological study design class was >80% . All thresholds reflect settings for ChromaTOF vs . 2 . 32 . Quantification was reported as peak height using the unique ion as default , unless a different quantification ion was manually set in the BinBase administration software BinView . A quantification report table was produced for all database entries that were positively detected in more than 10% of the samples of a study design class ( as defined in the miniX database ) for unidentified metabolites . A subsequent post-processing module was employed to automatically replace missing values from the * . cdf files . Replaced values were labeled as “low confidence” by color coding , and for each metabolite , the number of high-confidence peak detections was recorded as well as the ratio of the average height of replaced values to high-confidence peak detections . These ratios and numbers were used for manual curation of automatic report data sets to data sets released for submission . Metabolites were identified by matching the ion chromatographic retention index , accurate mass , and mass spectral fragmentation signatures with reference library entries created from authentic standard metabolites under the identical analytical procedure as the experimental samples . Relative levels of phosphorylated JNK1 and JNK3 normalized to GAPDH were analyzed using one-way ANOVA . Differences in JNK1 and JNK3 phosphorylation levels between SMI-treated and untreated mosquitoes and between transgenic and non-transgenic A . stephensi were determined by Student’s t-test . Survival analyses for JNK SMI-treated mosquitoes and controls and for transgenic and non-transgenic A . stephensi were performed using the Gehan-Breslow-Wilcoxon method . Analysis of median lifespans from independent replicates of JNK SMI-treated A . stephensi and MKP4 transgenic mosquitoes relative to controls was performed using one-way ANOVA . The proportions of females that laid eggs in the first gonotrophic cycle ( JNK SMI-treated versus controls ) were compared using Chi-square test , while egg clutch sizes per female in these studies were analyzed using Mann-Whitney test . Lifetime fecundity per female or for cages or populations of M3 and M4 lines of MKP4 transgenic A . stephensi relative to controls were calculated using Student’s t-test . Infection data were initially analyzed by ANOVA to determine whether mean oocysts ( for mosquitoes with at least one midgut oocyst ) differed among A . stephensi cohorts . No significant differences were found , so infection intensity data were analyzed using Kruskal-Wallis and Dunn’s post-test . Infection prevalences were analyzed by Chi-square test . Relative changes in P . falciparum growth were normalized to controls set at one and analyzed by Student’s t-test . mRNA expression data were analyzed by Student’s t-test to determine significance of changes relative to GAPDH ( for age-associated patterns ) or by one-way ANOVA for JNK SMI treatments and controls and for transgenic and non-transgenic A . stephensi . Luciferase activity levels of treatments and controls were analyzed by Student’s t-test . Fluorescent particle data for JNK SMI-treated A . stephensi and for transgenic and non-transgenic mosquitoes were analyzed by one-way ANOVA . All differences were considered statistically significant at α < 0 . 05 .
|
Malaria is a global health concern caused by infection with Plasmodium parasites . With rising insecticide and drug resistance , there is a critical need to develop novel control strategies . One strategy is to develop a Plasmodium-resistant mosquito through the manipulation of key signaling pathways and processes in the mosquito midgut , a critical tissue for parasite development . MAPK signaling pathways are highly conserved among eukaryotes and regulate development of the human malaria parasite Plasmodium falciparum in the mosquito vector . Here , we investigated how manipulation of Anopheles stephensi JNK signaling affects development of P . falciparum and key mosquito life history traits . We used multiple , complementary approaches to demonstrate that malaria parasite infection activates mosquito JNK signaling for its own benefit at a cost to host lifespan . Notably , these combined effects derive from networked signaling with other transduction pathways and alterations to intermediary metabolism in the mosquito host .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"phosphorylation",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"parasite",
"groups",
"body",
"fluids",
"plasmodium",
"parasitic",
"diseases",
"animals",
"parasitic",
"protozoans",
"parasitology",
"cell",
"signaling",
"c-jun",
"n-terminal",
"kinase",
"signaling",
"cascade",
"apicomplexa",
"protozoans",
"insect",
"vectors",
"infectious",
"diseases",
"malarial",
"parasites",
"proteins",
"metabolic",
"pathways",
"disease",
"vectors",
"insects",
"arthropoda",
"biochemistry",
"signal",
"transduction",
"mosquitoes",
"eukaryota",
"blood",
"anatomy",
"cell",
"biology",
"post-translational",
"modification",
"physiology",
"biology",
"and",
"life",
"sciences",
"species",
"interactions",
"metabolism",
"organisms",
"signaling",
"cascades"
] |
2018
|
Inhibition of JNK signaling in the Asian malaria vector Anopheles stephensi extends mosquito longevity and improves resistance to Plasmodium falciparum infection
|
The C . elegans ortholog of mammalian calsyntenins , CASY-1 , is an evolutionarily conserved type-I transmembrane protein that is highly enriched in the nervous system . Mammalian calsyntenins are strongly expressed at inhibitory synapses , but their role in synapse development and function is still elusive . Here , we report a crucial role for CASY-1 in regulating GABAergic synaptic transmission at the C . elegans neuromuscular junction ( NMJ ) . The shorter isoforms of CASY-1; CASY-1B and CASY-1C , express and function in GABA motor neurons where they regulate GABA neurotransmission . Using pharmacological , behavioral , electrophysiological , optogenetic and imaging approaches we establish that GABA release is compromised at the NMJ in casy-1 mutants . Further , we demonstrate that CASY-1 is required to modulate the transport of GABAergic synaptic vesicle ( SV ) precursors through a possible interaction with the SV motor protein , UNC-104/KIF1A . This study proposes a possible evolutionarily conserved model for the regulation of GABA synaptic functioning by calsyntenins .
A remarkable feature of the nervous system is the specific connections between neurons , which allows for orchestrated neural networks and circuitry of the brain . Many cell adhesion molecules ( CAMs ) are concentrated at synaptic sites in neuronal axons and dendrites , serving as dynamic regulators of synaptic function . Considering the complexity of the nervous system , tightly controlled spatial and temporal regulation of several different classes of CAMs is essential . Existing literature suggests that neuronal CAMs are not only important for adhesion but are also required for various aspects of synapse development and function [1–16] . The C . elegans casy-1 is an ortholog of mammalian Calsyntenin genes . Calsyntenins are type-I transmembrane proteins characterized by the presence of two cadherin-like tandem repeats , an LG/LNS domain in the extracellular region and an intracellular region that carries two kinesin light-chain binding domains [17 , 18] . All these regions are conserved in the three human Calsyntenin genes; clstn1 , clstn2 and clstn3 as well as in the sole C . elegans calsyntenin ortholog , casy-1 . Calsyntenins are highly expressed in the mammalian central nervous system [17 , 19] . Similarly , C . elegans CASY-1 was also observed in most head neurons and some other non-neuronal tissues like intestine and gonadal sheath cells [20 , 21] . Mammalian calsyntenins show a high degree of spatial and temporal regulation of expression , indicating possible diversity of functions . For example , calsyntenins associate with the scaffold proteins X11/X11L , which in turn associates with the amyloid precursor proteins ( APP ) forming a tripartite complex in the brain [22] . CLSTN1 has been shown to interact with the kinesin light-chain protein resulting in blocking the transport of APP-containing vesicles and thus causing more β- amyloid generation , suggesting a possible role of these molecules in the pathogenesis of Alzheimer’s disease [18 , 23–27] . Developmentally , CLSTN1 has been shown to be required for trafficking of NMDA receptors at the synapse and is essential for neuronal maturation during early embryonic development [28] . Further , CLSTN1 also regulates axon branching and endosomal trafficking during sensory neuron formation [29] . More recently , CLSTN1 has been shown to mediate trafficking of axon guidance receptors at spinal cord choice points [30] and in regulating microtubule polarity during sensory axon arbor development [31] . In contrast to the roles of CLSTN1 , which appears to mediate trafficking functions during nervous system development , CLSTN2 has been reported to be required for modulating synaptic plasticity . In a genome-wide screen for human hippocampus based-episodic memory genes , CLSTN2 was identified as a target that could be involved in human memory performance [32] . CLSTN2 knock-out mice are hyperactive with cognitive deficits due to reduced GABAergic neurotransmission [33] . The molecular basis of how CLSTN2 regulates GABA neurotransmission is however still unknown . CLSTN3 has been reported to act as a synaptic adhesion molecule that promotes excitatory and inhibitory synapse development in concert with neurexins [34] . In Zebrafish , the extracellular cadherin domains of calsyntenins have been shown to mediate homophilic adhesion [35] . However , despite all these studies that have allowed us to understand the functions of calsyntenins , we are still far from elucidating the molecular and physiological underpinnings of this molecule . The calsyntenin ortholog in C . elegans , CASY-1 has also been found to be essential for multiple forms of learning [20 , 21] . It has been shown that in salt chemotaxis learning , the LG/LNS domain in the extracellular region of CASY-1 is essential for memory formation . Functional rescue experiments by expressing human CLSTN2 in C . elegans casy-1 mutants were able to rescue the learning defects , highlighting conservation of function . In this study , we propose a role for CASY-1 in regulating GABA synaptic vesicle ( SV ) precursor transport and hence maintaining normal GABAergic neurotransmission at the C . elegans neuromuscular junction ( NMJ ) . We show that the two shorter isoforms of CASY-1 , that only have the conserved C-terminal region and lack the entire extracellular N-terminal region , function in GABAergic motor neurons to regulate GABA release by mediating the trafficking of GABA SV precursors via their interaction with the UNC-104 motor protein .
The C . elegans NMJ is an extensively used model to understand various aspects of synapse development and function . The NMJ consists of body-wall muscles that receive synaptic inputs from both excitatory cholinergic and inhibitory GABAergic motor neurons . An intricate stability between the excitatory and inhibitory signaling is responsible for the sinusoidal locomotion in C . elegans and any defect in this balance could result in altered synaptic function ( reviewed in [36 , 37] ) . Previously an RNAi screen was conducted using the acetylcholine esterase inhibitor , Aldicarb , to identify the function of cell adhesion molecules at the C . elegans NMJ [38] . The presence of Aldicarb causes acute paralysis due to accumulation of acetylcholine at the NMJ . Loss of function of genes that are necessary for synaptic function could cause either increased resistance or hypersensitivity to Aldicarb [39–42] . One of the positives from the screen was casy-1 . To validate the results of RNAi screening , two different mutant alleles of casy-1 , tm718 and hd41 , were obtained . Both alleles are putative null alleles as they carry deletions that start at the N- terminal region and result in frame-shift mutations ( [21] and ( illustrated in Fig 1A ) ) . Both casy-1 ( tm718 ) and casy-1 ( hd41 ) showed significant hypersensitivity to Aldicarb suggesting neuromuscular signaling defects in these mutants ( Fig 1B ) . All further experiments were performed using the casy-1 ( tm718 ) mutant allele . The casy-1 locus in C . elegans is predicted to encode three isoforms based on EST evidence [43] . CASY-1A , a 984 residue full-length protein contains all the conserved domains of mammalian calsyntenins ( illustrated in Fig 1A ) . CASY-1B and CASY-1C are truncated proteins encoding 167 and 160 residues respectively and lack most of the N- terminal of the calsyntenin gene ( illustrated in S1A Fig ) . Real-time qPCR experiments revealed that all three isoforms of casy-1 are significantly reduced in the casy-1 ( tm718 ) mutant allele ( S1B Fig ) . An isoform-specific rescue of the casy-1 mutant phenotype was then performed using each isoform with its native promoter ( illustrated in Fig 1D and detailed in supplemental methods ) . Surprisingly , all three isoforms could fully rescue the Aldicarb hypersensitivity of casy-1 mutants ( Fig 1C ) suggesting that ( 1 ) the promoter sequences used for each casy-1 isoform appears to be functional and ( 2 ) all three isoforms could be functioning to regulate synaptic transmission at the NMJ . As reported earlier [21] , we observed strong expression of casy-1a in a lot of head neurons including the amphid sensory neurons , the Ventral Nerve Cord ( VNC ) and some tail neurons . Expression was also observed in somatic tissues like gonadal sheath and intestine ( Fig 1E ) . To examine the expression pattern of casy-1b and casy-1c , transcriptional reporter lines expressing NLS-GFP under isoform-specific promoters were generated . Compared to casy-1a , which is highly enriched in the head and tail neurons , casy-1b and casy-1c show a more limited expression in the head neurons . However , both casy-1b and casy-1c are strongly expressed in the VNC motor neurons , which is not seen with casy-1a . The weak expression of casy-1a in the VNC belongs to the axonal processes from the head neurons that traverse along the entire length of the C . elegans body . No expression was observed in the gonadal sheath for the shorter isoforms ( Fig 1E ) . Mammalian orthologs of CASY-1 have been shown to regulate various aspects of neuronal development [28–31 , 34] . Since defects in the synthesis or release of either excitatory ( acetylcholine ) or inhibitory ( GABA ) neurotransmitter could result in Aldicarb hypersensitivity [41] , we speculated that defects in cholinergic or GABAergic neuronal development could explain the Aldicarb hypersensitivity in casy-1 mutants . To examine this , transgenic lines that express soluble markers driven by acetylcholine ( ACh; unc-17 ) or GABA ( unc-25 ) neuron-specific promoters were used . These transgenes were introduced in the casy-1 mutant background . Analysis of the imaging data revealed no gross morphological defects in GABAergic or cholinergic motor neurons in casy-1 mutants ( Fig 2A and S2A Fig ) . To further validate that CASY-1 is not essential for motor neuron developmental functions , a transgenic line that expressed the CASY-1A isoform conditionally under a heat shock promoter hsp16 . 2 was used . Aldicarb hypersensitivity of casy-1 mutants was completely rescued , and the C . elegans became resistant to Aldicarb when CASY-1A was expressed transiently in the mature nervous system ( Fig 2B ) . These results further indicate the role of CASY-1 in regulating synaptic function in a mature nervous system rather than in neuronal development . Since all three CASY-1 isoforms can completely rescue the Aldicarb hypersensitivity of casy-1 mutants ( Fig 1C ) , we decided to determine the region of CASY-1 that is required to regulate synaptic transmission at the NMJ . Two transgenic lines were utilized in which either the entire extracellular N-terminal region [CASY-1A ( ΔN ) ] or the entire intracellular C-terminal region [CASY-1A ( ΔC ) ] were removed [21] . Both constructs were expressed under the ins-1 promoter in the ventral cord motor neurons [44] . Transgenes lacking the N-terminal domains could rescue the Aldicarb hypersensitivity of casy-1 mutants , however transgenes lacking the C-terminal region could not rescue the Aldicarb defects ( Fig 2C ) . These results indicate that the extracellular domains are dispensable for the synaptic function of CASY-1 at the NMJ . The requirement of just the cytoplasmic C-terminal region for regulating synaptic function at the NMJ further explains how all three CASY-1 isoforms , which are structurally different could rescue the Aldicarb defect of casy-1 mutants . Since the entire C-terminus is conserved in all three CASY-1 isoforms ( S2B Fig ) it is conceivable that they utilize this region for their synaptic function when expressed in motor neurons . To identify specific neurons where CASY-1 could be functioning to regulate synaptic transmission , all isoforms of CASY-1 were expressed under cholinergic ( unc-17 ) or GABAergic ( unc-25 ) neuron-specific promoters . All three isoforms completely rescued the Aldicarb hypersensitivity of casy-1 mutants when expressed in GABAergic neurons but not in cholinergic neurons ( Fig 3A ) . These results indicate that CASY-1 could be regulating synaptic transmission in GABA motor neurons at the NMJ . Although expression of CASY-1 isoforms was not observed in the body-wall muscle , however , to remove the possibility of any retrograde signaling from the body-wall muscle effecting neuronal behavior at the NMJ , a transgenic line expressing CASY-1A under a body-wall muscle specific promoter ( myo-3 ) was generated . Expression of CASY-1A in body-wall muscles failed to rescue the Aldicarb hypersensitivity of casy-1 mutants ( Fig 3A ) . Previous experiments ( Fig 2C ) show that the C-terminal of CASY-1 is sufficient to rescue the Aldicarb hypersensitivity of casy-1 mutants . To further strengthen our hypothesis , a transgenic line CASY-1A ( ΔC ) was generated , in which the entire C- terminal ( 880–984 aa ) was removed from CASY-1A and expressed under the GABA-specific promoter ( unc-25 ) . This line failed to rescue the Aldicarb defects of casy-1 mutants ( Fig 3B ) , further signifying the importance of the C-terminal of CASY-1 at the NMJ . To further investigate the role of GABA signaling in casy-1 mutants , an assay using the drug Pentylenetetrazole ( PTZ ) , a potent antagonist of GABAA receptors [45] , was performed . PTZ has been shown to be effective in C . elegans for generating a convulsive phenotype called ‘head bobs’ , which is used as an indicator of reduced GABAergic synaptic-transmission [46] . Mutants in unc-25 , a GABA-synthesizing enzyme , were used as positive controls ( S3A Fig ) . The casy-1 mutant animals showed a strong convulsion phenotype after a 30-minute exposure to 10mg/ml PTZ and by the end of 60 minutes , a significant fraction of casy-1 mutants showed convulsions ( Fig 3C ) . Additionally , casy-1 mutants also showed a tail shrinking phenotype , a characteristic of GABA mutants in C . elegans [47] ( S1–S3 Movies ) . The convulsive phenotype of casy-1 mutants was completely rescued by expressing the CASY-1 isoforms in GABAergic neurons but not in cholinergic neurons or muscle ( Fig 3C ) further implicating the role of CASY-1 in GABA signaling . Reduced GABAergic synaptic transmission could be either due to reduced release of GABA from the GABAergic motor neurons ( pre-synaptic ) or reduced response of muscle to GABA due to lower expression of GABAA receptors on the muscle ( post-synaptic ) . To differentiate between these two possibilities , pre-synaptic and post-synaptic markers were analyzed in the casy-1 mutant background . Initially the expression of pre-synaptic markers under cholinergic and GABAergic neuron specific promoters were investigated . First , GFP–tagged SNB-1 , a C . elegans ortholog of the mammalian Synaptic Vesicle ( SV ) protein synaptobrevin was analyzed . Synaptobrevin localizes in a punctate pattern along the dorsal cord of C . elegans [40] . Mutants in casy-1 showed a subtle but significant decrease in GABAergic::SNB-1::GFP levels , suggesting fewer GABA vesicles at the synapse ( Fig 3D ) . Analysis of the Dorsal Nerve Cord ( DNC ) puncta to VNC cell body fluorescence ratio further displayed a significant decrease supporting the presence of fewer GABA vesicles at the DNC synapses ( S3B Fig ) . Fluorescence intensity of cholinergic synapses was indistinguishable in casy-1 mutants from wild-type ( WT ) animals , suggesting normal cholinergic signaling ( Fig 3E ) . The synapse density in the dorsal cord was also examined , by quantifying SYD-2 puncta , an ortholog of the mammalian active zone protein α-Liprin [40] . The density of cholinergic and GABAergic synapses were largely normal , indicating that synapse development is largely unaffected in casy-1 mutants ( S3C and S3D Fig ) . The C . elegans body muscles express two classes of Acetylcholine receptors ( AChRs ) ; nicotine-sensitive ( nAChR ) and levamisole-sensitive ( LAChR ) receptors as well as a single class of GABAA receptors . There was no significant change in the fluorescent intensity of the GFP- tagged cholinergic receptors ( S3E Fig ) in the mutants . The GABAA receptor , UNC-49 also showed similar levels of expression and localization in casy-1 mutants when compared to WT animals ( S3F Fig ) . To further ensure that post-synaptic GABAA receptors functioned normally in the mutants , a GABA agonist Muscimol was utilized . C . elegans placed on Muscimol show various degrees of responses depending on their sensitivity to the drug . The most characteristic phenotype is the “rubber band phenotype” , wherein the animals contract and relax without displacement following prodding on the head [48] . If reduced GABAergic signaling in casy-1 mutants is due to lower release of GABA from the pre-synaptic terminal , no difference in the severity of the rubber band phenotype would be observed following incubation with Muscimol . When casy-1 mutants were tested in the Muscimol assay they behaved like WT C . elegans ( S3G Fig ) , again strengthening our hypothesis that CASY-1 functions pre-synaptically at the NMJ . To further ensure that cholinergic transmission is normal in casy-1 mutants , we performed a Levamisole assay . Levamisole is an agonist for L-type cholinergic receptors and exposure to Levamisole results in a time-course induced paralysis in WT animals . Mutants in casy-1 showed a response similar to WT animals in the Levamisole assay suggesting normal LAChR signaling at the NMJ ( S3H Fig ) . The C . elegans nervous system consists of 26 GABAergic neurons which include six DD and thirteen VD neurons that innervate the dorsal and ventral body-wall muscles respectively , four RME motor neurons that innervate head muscles , the AVL and DVB neurons that are pre-synaptic to the enteric muscles and RIS which is an interneuron [47 , 49] . To confirm that CASY-1 functions specifically in GABA motor neurons ( D- type neurons ) , a transgenic line expressing the CASY-1C isoform specifically in DD and VD class of GABA motor neurons using the unc-30 promoter that also expresses in some non-GABAergic neurons was generated [50] . Expression of CASY-1C in just D-type motor neurons completely rescued the Aldicarb hypersensitivity of casy-1 mutants , indicating that CASY-1 functions in motor neurons to regulate synaptic transmission at the NMJ ( Fig 4A ) . To further explore the function of casy-1 in GABA neurotransmission , we utilized an optogenetic tool , where channelrhodopsin ( ChR2 ) , a light-gated cation channel that allows non-specific flow of cations into the cell leading to electrical excitation , was used [51] . A transgenic line that expressed ChR2 specifically in GABAergic neurons was utilized . The light-based activation of GABA motor neurons leads to a measurable change in body length ( relaxation ) . Activation of GABAergic ChR2 results in a reduced relaxation in casy-1 mutants in comparison to WT controls . These results suggest a lower release of GABA from the pre-synaptic termini in casy-1 mutants , which in turn results in less relaxation upon optogenetic stimulation . Expressing the CASY-1C isoform in GABAergic neurons completely rescued the decreased relaxation defect of casy-1 mutants ( Fig 4B ) . To evaluate changes in endogenous synaptic transmission , whole-cell patch clamp recordings of the muscles under voltage-clamp conditions from dissected C . elegans were analyzed . Endogenous excitatory and inhibitory post-synaptic currents ( EPSCs and IPSCs ) that indicate the frequency and amplitude of neurotransmitter release from cholinergic and GABAergic motor neurons onto the body-wall muscle were measured in the WT and casy-1 mutant animals . The casy-1 mutants had a significantly lower endogenous IPSC rate compared to WT animals ( Fig 4C ) . The amplitude of IPSCs was unaltered . Collectively , these results indicate that the release of GABA is reduced in casy-1 mutants while the muscle responsiveness to GABA remained unaffected . Expression of CASY-1C specifically in GABA motor neurons significantly rescued the decreased IPSC frequency in the casy-1 mutants ( Fig 4C ) . To validate the specific role of shorter casy-1 isoforms in regulating GABA signaling , a transgenic line that expresses CASY-1A isoform under its endogenous promoter was tested to rescue the decreased IPSC frequency in the casy-1 mutants . This transgenic line did not show significant differences in IPSC rate when compared to the casy-1 mutant C . elegans . This experiment further strengthens our data that the shorter isoforms expressed in motor neurons function to modulate GABAergic transmission ( S4 Fig ) . Interestingly , we found that the endogenous EPSC rate showed a significant increase in casy-1 mutants when compared to WT animals ( Fig 4D ) . Expression of CASY-1C specifically in GABA motor neurons could not rescue the EPSC frequency in the casy-1 mutants ( Fig 4D ) . The amplitude of EPSCs was again unaffected . Changes in acetylcholine release from cholinergic motor neurons are likely to be secondary , as functional rescue experiments showed that expressing casy-1 in GABAergic , but not in cholinergic motor neurons , rescue the Aldicarb hypersensitivity of casy-1 mutants . Thus , it is possible that the increased EPSC frequency in casy-1 mutants could be due to the function of CASY-1 in higher levels of neuronal circuits ( e . g . sensory or interneurons ) . Expression of all CASY-1 isoforms in GABAergic neurons completely rescued the Aldicarb hypersensitivity of casy-1 mutants ( Fig 3A ) . Also , casy-1b and casy-1c transcriptional reporters express in motor neurons ( Fig 1E ) . To determine the specific class of motor neurons where casy-1 isoforms are expressing , isoform specific- GFP transcriptional reporters were co-expressed with cholinergic or GABA specific mCherry-tagged reporter lines . The casy-1a transcriptional reporter showed no co-localization with either cholinergic or GABAergic motor neurons ( Fig 5A , column one ) . However , casy-1b and casy-1c showed co-localization with both cholinergic and GABAergic neuronal markers ( Fig 5A , columns two and three ) . This data suggests that the shorter isoforms of casy-1 are probably functioning in GABA motor neurons to regulate GABA release and hence affect synaptic transmission at the NMJ , however , the role of these isoforms in cholinergic neurons is still unclear . To examine specific subcellular localization of the shorter casy-1 isoforms , CASY-1C::mCherry was expressed under its own promoter . The CASY-1C::mCherry protein localizes all along the dorsal cord axons and shows co-localization with the GABAergic and cholinergic pre-synaptic SNB-1::GFP markers ( Fig 5B and 5C ) . This suggests the presence of CASY-1 at the synapse . The expression of CASY-1C::mCherry at the DNC synapses was diffuse rather than punctate . This could be due to the lateral diffusion followed by ecto-domain shedding at the synaptic cleft ( described below ) . Expression of mCherry and GFP- tagged CASY-1C fully rescued the Aldicarb sensitivity of casy-1 mutants suggesting that the tagged proteins are functional ( S5A Fig ) . Mammalian calsyntenins have been shown to be cleaved at their extracellular region [19 , 23] . Further , the CASY-1A isoform has previously been shown to be cleaved in its extracellular region juxtaposed to the membrane by synaptic cleft peptidases , resulting in the release of the entire N- terminal into the synaptic cleft [21] ( illustrated in S5B Fig ) . To figure out if CASY-1B and CASY-1C are also cleaved once they reach the synapse , N- terminal mCherry fusion transgenes of CASY-1B and CASY-1C were generated . The mCherry tagged N-terminal of the shorter CASY-1 isoforms was detected in the coelomocytes ( S5B Fig; bottom panels ) , which are macrophage- like scavenger cells that take up any waste or secreted proteins from the body cavity ( reviewed in[52] ) . These results allow us to conclude that all CASY-1 isoforms are cleaved at their N-terminal region by synaptic cleft peptidases , resulting in the release of the ectodomain into the body cavity . These results also provide appropriate justification to how CASY-1A isoform , which does not express in motor neurons could rescue the Aldicarb hypersensitivity when expressed in GABA motor neurons . CASY-1 isoforms are present on the SV precursors with their N-terminal facing the SV lumen , while the conserved C-terminal faces towards the cytosol where it can interact with a wide variety of cytoplasmic proteins . Since the entire C-terminal is totally conserved in the three casy-1 isoforms , expression of any of the three isoforms in GABA motor neurons could potentially rescue the Aldicarb phenotype of the casy-1 mutants . Although this data could explain the rescue of Aldicarb hypersensitivity by the CASY-1A isoform in GABA motor neurons , the question still remains as to how the expression of CASY-1A under its endogenous promoter could rescue the Aldicarb hypersensitivity in casy-1 mutants ( Fig 1C ) . CASY-1A expression is highly enriched in the head sensory neurons . Previous reports have highlighted the role of genes that function in higher levels of locomotion circuit to regulate motor circuit activity at the NMJ [53 , 54] . Future investigations examining CASY-1A functions in sensory neurons might shed some light on this aspect . Reduced SNB-1::GFP levels , less relaxation upon optogenetic stimulation of GABAergic neurons and decreased endogenous IPSC rate ( Figs 3D , 4B and 4C ) all highlight the role of CASY-1 in regulating GABA release from motor neurons . To further address this role of casy-1 , a transgenic line in which the luminal domain of synaptobrevin is tagged with superecliptic pHluorin; a GFP reporter expressed specifically in GABAergic neurons [55] was utilized . pHluorin is highly pH-sensitive and its fluorescence remains quenched in the acidic environment of the SV lumen , however , there is a dramatic increase in the fluorescence as soon as the vesicle fuses onto the membrane relieving the tag from the acidic environment of the SV [56 , 57] . To examine if casy-1 mutants have fewer GABA vesicles at the synapse , and hence less release of GABA , we monitored the fluorescence intensity of pHluorin at the dorsal cord synapses . pHluorin intensity was significantly reduced in the dorsal cord synapses of casy-1 mutants further confirming the role of CASY-1 in the release of GABA at NMJ ( Fig 6A ) . Further , we addressed the possibility of the involvement of CASY-1 in GABA vesicle trafficking by performing fluorescence recovery after photobleaching ( FRAP ) analysis of SNB-1::GFP levels . The recovery rate of SNB-1::GFP depends upon two factors; transport of new SV precursors at the synapse by motor-mediated trafficking or diffusion from neighboring synapses into the bleached area . The recovery rate of GABAergic SNB-1::GFP was significantly reduced in casy-1 mutants suggesting that the mobility dynamics of SVs at the NMJ is compromised in the mutants . The recovery rate was completely restored by expressing the CASY-1C isoform specifically in GABAergic neurons ( Fig 6C ) . We also monitored the mobility dynamics in cholinergic motor neurons but found no significant change in casy-1 mutants compared to WT animals ( Figs 6C and S6 ) , suggesting that CASY-1 is specifically functioning to regulate SV release in GABA motor neurons . In C . elegans , the motor neuron soma that are present along the ventral nerve cord send out axons in the anterior direction , which extend commissures dorsally by crossing the midline and fasciculate with the DNC to form synapses with the body wall muscles . To address how casy-1b/casy-1c regulate the trafficking of SV precursors from motor neuron cell bodies to the DNC synapses , the transport characteristics of SNB-1::GFP tagged vesicles along the commissures were examined using time lapse imaging . We first assayed SV transport in D- type GABAergic motor neurons . In WT animals , SNB-1::GFP labeled vesicles are transported in both anterograde and retrograde directions , with an anterograde bias ( Fig 7B ) . However , in casy-1 mutants , the SV anterograde transport was significantly reduced , although , the velocity of SNB-1::GFP remains similar to WT animals . Decrease in anterograde vesicular flux was rescued by expressing the CASY-1C isoform in GABAergic motor neurons ( Fig 7A–7C and S4–S6 Movies ) . We next analyzed the transport characteristics of SNB-1::GFP vesicles in Cholinergic motor neurons . Vesicular velocity for SNB-1::GFP was higher in Cholinergic neurons compared to GABAergic neurons . Analysis of SNB-1::GFP transport in cholinergic motor neurons of casy-1 mutants showed that anterograde and retrograde vesicular flux and velocity are similar to that of wild type , suggesting that vesicular transport is not affected in cholinergic neurons in casy-1 mutants . ( Fig 7D–7F and S7 and S8 Movies ) . This data suggests that the shorter CASY-1 isoforms function specifically in GABAergic motor neurons to modulate the release kinetics of GABA by directly influencing the transport of SV precursors . The expression of different vesicular cargo proteins in the casy-1 mutants was next examined in order to investigate if casy-1 mutants have general trafficking defects . The reduced SNB-1::GFP levels in the GABA motor neurons ( Fig 3D ) were due to decreased transport of GABA SV precursors at the synapse and hence suggests that monitoring the fluorescent intensity of tagged cargoes might provide an indication of possible trafficking defects in the casy-1 mutants . The transport characteristics of cholinergic SNB-1::GFP were unaltered in the casy-1 mutants which clearly suggests that defects in vesicular cargo trafficking is not a general phenomenon in casy-1 mutants . To validate this further , the fluorescence intensity of diverse vesicular cargo was monitored at the GABAergic DNC synapses in casy-1 mutants . The fluorescence intensity of tagged- mitochondrial cargo ( MITO::GFP ) as well as early endosomes ( RAB-5::YFP ) was not affected in the GABAergic neurons of casy-1 mutants ( S7A and S7B Fig ) . However , the fluorescence intensity of a lysosomal marker ( CTNS-1::GFP ) was significantly reduced suggesting the presence of fewer lysosomes at the synapse ( S7C Fig ) . This decrease in lysosomal marker fluorescence could be due to absence of the interaction of CASY-1 with the kinesin light chain motor proteins like klc-2 that forms functional complexes with the kinesin heavy chain motor proteins unc-116 required for trafficking of lysosomal cargos [58 , 59] . Mutants in casy-1 have fewer GABA vesicles at the synapse as demonstrated by a significant decrease in SNB-1::GFP levels at the NMJ ( Fig 3D ) . They also show reduced recovery rates of SNB-1::GFP following FRAP analysis , suggesting the possibility that fewer SV precursors are trafficked to the NMJ ( Fig 6B ) . UNC-104/KIF1A is a kinesin-3 motor that has a conserved role in trafficking SV precursors from the cell body to the synapse [60–63] . To address the possibility that CASY-1 could be interacting with UNC-104 to regulate the anterograde trafficking of SVs in GABAergic motor neurons , Aldicarb assays were performed after silencing the casy-1 gene in the unc-104 mutant background using RNAi . The reverse experiment of knocking down unc-104 in the casy-1 mutant background was also performed . We performed RNAi knockdown experiments as we were unable to make unc-104; casy-1 double mutants due to close proximity of the genes on Chromosome II . Both the knock-down and knock-out of unc-104 , resulted in resistance to Aldicarb as has been previously reported [64] . Further , RNAi knock-down of unc-104 in the casy-1 mutant background , completely abolished the hypersensitivity of casy-1 mutants and resulted in resistance to Aldicarb just like what was seen in unc-104 mutants , suggesting that UNC-104 may genetically interact with CASY-1 ( S8A Fig ) . However , resistance to Aldicarb could also be due to the dominant phenotype of unc-104 over casy-1 gene function . To further validate the genetic interaction between casy-1 and unc-104 , FRAP analysis of SNB-1::GFP in GABAergic motor neurons after RNAi knockdown of unc-104 in a casy-1 mutant background was performed . The recovery rate of SNB-1::GFP was significantly reduced after knockdown of unc-104 . However , no significant difference between the recovery rates was observed after unc-104 knockdown in WT and casy-1 mutant background , further supporting a genetic interaction between casy-1 and unc-104 ( Fig 8A ) . The C . elegans CASY-1 has previously been shown to physically interact with the kinesin light chain-2 ( KLC-2 ) in yeast two-hybrid assays [59] . These studies suggest CASY-1 acts as a broad regulator for transport of multiple neuronal cargoes . To investigate the role of CASY-1 in general transport mechanisms , a possible physical interaction between CASY-1 and UNC-104 in a yeast two-hybrid assay was examined . The interaction of the CASY-1 C- terminal with several UNC-104 domain constructs was observed . Our data suggests that a weak interaction occurs between the cytoplasmic tail of CASY-1 and the C-terminal of UNC-104 that largely includes the stalk region and the Pleckstrin Homology ( PH ) domain , a domain responsible for cargo vesicle binding to UNC-104 [65] . Subtle interactions were also observed with the UNC-104 motor domain and FHA domain . This data suggests that the C-terminal of CASY-1 could directly interact with UNC-104 . ( S8B Fig ) . However , since the interaction between UNC-104 and CASY-1 appeared weak in the yeast-two-hybrid assay , we could not rule out the possibility that CASY-1 interacts indirectly with UNC-104 , through other adaptor molecule/s . After identifying the C-terminal PH domains and adjacent regions as critical mediators of CASY-1 and UNC-104 interactions , a GST pull down experiment was performed to get better insights into the physical interaction between these proteins . The C- terminal region of C . elegans CASY-1C and the C- terminal without the kif interacting domain , CASY-1C ( ΔKIF ) , were expressed as GST fusion proteins in bacteria . CASY-1C ( ΔKIF ) represents a deletion in the region harboring the two WDDS motifs and the acidic region ( amino acids 70–148 ) in the C- terminal region of CASY-1C ( deletion schematized in Fig 8B , top panel ) . GST alone , GST fused with CASY-1C and CASY-1C ( ΔKIF ) were used to precipitate HA- tagged UNC-104 protein expressed in a bacterial lysate . Full length CASY-1C significantly precipitated the 110 KDa C-terminal region of HA-tagged UNC-104 ( Fig 8B ) . However , CASY-1C ( ΔKIF ) GST fusion did not show any precipitation of the HA-tagged UNC-104 ( Fig 8C ) . This data further suggests that the C- terminal of CASY-1C interacts with the tail region of UNC-104 . Additionally , UNC-104 and CASY-1C interaction was further validated using bimolecular fluorescence complementation ( BiFC ) assays , a method widely used to study protein-protein interactions in live animals . Here UNC-104 and CASY-1C were fused to a non-fluorescent YFP ( Venus ) hybrid also called ‘split-YFP’ ( made up of the VN or VC of fluorescent YFP signals ) . Expression of YFP indicates that these two proteins are closely located ( less than 7 nm apart ) thus allowing for the fluorophore complementation to occur , hence leading to visible fluorescence [66–69] . As a positive control , an Punc-104::UNC-104::VN/ Punc-104::UNC-104::VC BiFC pair was microinjected as it has been well established that monomeric UNC-104 dimerizes to become functional and this dimerization can be tested using BiFC ( [69] and Fig 8C ) . As a negative control , an Punc-104::UNC-104::VN/ empty::VC BiFC pair was examined . No fluorescence was detected in these transgenic C . elegans . The experimental transgenic line containing Punc-104::UNC-104::VN/ Punc-25::CASY-1C::VC showed significant YFP signal patterns along the GABAergic motor neurons on the VNC ( Fig 8C ) . YFP fluorescence signals were significantly reduced in the transgenic animals containing Punc-104::UNC-104::VN/ Punc-25::CASY-1C ( ΔKIF ) ::VC BiFC pair ( Fig 8C ) . Finally we found that the expression of CASY-1C ( ΔKIF ) in GABAergic motor neurons could not rescue the Aldicarb phenotype in casy-1 mutants ( S8C Fig ) . These data strongly support functional in vivo interactions between UNC-104 and CASY-1C . To further understand the role of UNC-104 in axonal trafficking of CASY-1 , we determined the localization of CASY-1C::GFP in unc-104 mutant animals . In WT C . elegans , CASY-1C::GFP shows significantly lower expression in the VNC cell bodies owing to its trafficking to the DNC synapses . However , in unc-104 mutants we observed significant accumulation of CASY-1C::GFP in some neuronal cell bodies ( Fig 8D ) . In parallel , we observed a significant decrease in the fluorescence intensity of CASY-1C::GFP at the DNC synapses ( Fig 8E ) , suggesting that UNC-104 is a potential candidate for trafficking of CASY-1C::GFP to the synapse in these neurons . We next examined the motor neurons where CASY-1C::GFP is accumulating in the unc-104 mutant animals . The localization pattern of Punc-25::mCherry with CASY-1C::GFP in the unc-104 mutants revealed that CASY-1C accumulation in the unc-104 mutants is not specific to GABAergic motor neurons as significant accumulation was observed in the adjacent cells that represent cholinergic motor neurons ( arrow in S8D Fig ) . Together , our results strongly suggests an interaction between the C-terminal of the shorter CASY-1 isoforms with the UNC-104 motor protein that appears to facilitate the transport of GABA SV precursors required to maintain inhibitory GABAergic signaling at the NMJ .
GABA is a major inhibitory neurotransmitter functioning in both vertebrate and invertebrate nervous systems . In vertebrates , nearly 30–40% of the CNS synapses are thought to be GABAergic [70] and alterations in GABA neurotransmission have been associated with several different neurological disorders ( reviewed in [71–73] ) . In C . elegans , very few reports have documented the role of genes regulating GABAergic synaptic transmission at the NMJ [74 , 75] . In this study , we are proposing that the shorter CASY-1 isoforms are essential for regulating GABA signaling at the NMJ . Despite considerable reports emphasizing the involvement of calsyntenins in GABA synapse development and function , it has been very difficult to deduce the cellular and molecular mechanisms and implications of these proteins on animal behavior and function . Our results indicate that the short isoforms of CASY-1 , CASY-1B and CASY-1C , are involved in modulating GABA synaptic release from motor neurons via interaction with the UNC-104 motor protein ( Illustrated in Fig 9 ) . Mutants in casy-1 show significant hypersensitivity to Aldicarb , suggesting increased synaptic transmission at the NMJ . Neuron-specific rescue experiments suggested that the shorter CASY-1 isoforms function specifically in GABAergic neurons to regulate synaptic signaling . Although the Aldicarb assay is a direct assay for monitoring defects in cholinergic signaling , it is also routinely used to identify mutants defective in GABAergic synaptic transmission . The increased hypersensitivity in Aldicarb assay , could result either from increased acetylcholine release ( presynaptic defect ) or increased expression of cholinergic receptors on the muscle ( postsynaptic defect ) . Aldicarb hypersensitivity could also result from decreased GABA release ( presynaptic defect ) or decreased expression of GABA receptors ( postsynaptic defect ) [41 , 76 , 77] . Since motor circuit activity results from a balance between excitatory and inhibitory signaling at the NMJ , any defect in GABA signaling will result in a change in excitatory to inhibitory ratio , giving an Aldicarb phenotype [78–81] . Mutants in GABA signaling like unc-25 ( GABA-synthesis enzyme ) and unc-47 ( GABA transporter ) have been reported to be Aldicarb hypersensitive , due to overall increase in excitatory cholinergic signaling [41 , 82 , 83] . Our domain-mapping experiments suggest that the conserved C-terminal of CASY-1 , present in casy-1b/casy-1c is functioning to rescue the hypersensitive phenotype of casy-1 mutants . The expression analysis of CASY-1 isoforms further showed that only CASY-1B and CASY-1C express in GABAergic motor neurons , while CASY-1A expression is restricted mainly to head neurons and is not seen in the ventral cord motor neurons . Interestingly , C . elegans has devised a fascinating strategy wherein isoform expression and function is spatially regulated using alternative promoters . The shorter isoforms of CASY-1 , which are essentially just the C-terminal region of mammalian Calsyntenins and required for this function , are expressed in GABAergic motor neurons to regulate GABA release at NMJ . As discussed before , several studies of mammalian synapses using primary hippocampal neuronal cultures and knockout mice have established that calsyntenins are involved in the development and functioning of inhibitory GABA synapses [34] . CLSTN knockout mice also show reduced GABAergic neurotransmission [33] but the molecular basis of this regulation is unknown . In our study , we are showing , how the C-terminal of mammalian Calsyntenins which is conserved in CASY-1B/CASY-1C , can regulate GABAergic neurotransmission pre-synaptically . Using pharmacological , behavioral , optogenetic and electrophysiology approaches , we established defects in GABA signaling in casy-1 mutants at the NMJ . All these mutant phenotypes could be completely rescued by expressing CASY-1B/CASY-1C specifically in GABAergic motor neurons . Although , mammalian calsyntenins are reported to be post-synaptic membrane proteins , here we are demonstrating a pre-synaptic role of the C . elegans orthologs of mammalian Calsyntenins . Our study opens up the possibility of exploring the potential existence of similar mechanisms regulating GABAergic neurotransmission in the mammalian nervous system . We are also throwing light into the mechanistic insight of how the CASY-1 C-terminal could regulate GABA release at NMJ . Mammalian CLSTN1 has been well documented for its role in regulating trafficking of various axo-dendritic synaptic components via its interaction with kinesin light chain motor protein ( klc-1 ) . C . elegans CASY-1 has also recently been shown to interact physically with KLC-2 [59] , suggesting a role for CASY-1 in trafficking of synaptic components as well . In this study , we have determined a novel interaction where the C-terminal of CASY-1 interacts with the tail region of the motor protein UNC-104 that essentially harbors the stalk region as well as the PH domain . UNC-104/KIF1A is established as an evolutionarily conserved motor protein required for trafficking of SV precursors from the soma to the synapse . Our results unlock the possible existence of similar interaction in mammalian system . However , we cannot nullify the possibility that CASY-1 might also affect SV precursor trafficking via its interaction with kinesin light chain ( klc-1 and klc-2 ) as multiple studies also showed the involvement of klc-1 motor in SV precursor trafficking [84–86] . Despite proposing a convincing mechanistic role of casy-1 isoforms in modulating GABA signaling at NMJ , some major questions remain unanswered . First , casy-1b/casy-1c specifically act in GABAergic neurons although they are also expressed in cholinergic neurons . Several previous reports have highlighted the role of molecules that expresses in both cholinergic and GABAergic motor neurons but specifically functions in just one system to affect NMJ signaling [87 , 88] . Here we establish that casy-1 isoforms function in GABAergic motor neurons to regulate Aldicarb responsiveness . However , we cannot rule out additional roles of CASY-1 in cholinergic motor neurons . Furthermore , electrophysiological recordings from the casy-1 NMJ showed an increased EPSC frequency suggesting another mechanism for Aldicarb hypersensitivity , but expressing CASY-1 isoforms in cholinergic motor neurons could not rescue the Aldicarb phenotype . This implies that increased EPSC frequency is not an outcome of casy-1 function in cholinergic motor neurons . Future investigations to address how CASY-1 could affect EPSC frequency might provide useful insights into other functions of CASY-1 isoforms that might affect neuromuscular signaling . Despite the simplicity of C . elegans having just one gene coding for CASY-1 , compared to multiple mammalian Calsyntenin genes , it emerges as an excellent regulator of diverse functions such as vesicular trafficking , functioning of GABAergic synapses and synaptic plasticity , functions that are performed by individual CLSTNs in the mammalian nervous system . This establishes C . elegans as an ideal model system to explore other functions of mammalian calsyntenins . Further studies in this system could enhance our understanding about pathophysiological mechanisms that trigger calsyntenin-related brain disorders .
All strains were maintained on nematode agar growth medium ( NGM ) plates seeded with OP50 Escherichia coli at 20°C under standard conditions [89] . The C . elegans Bristol strain , N2 was used as the wild-type ( WT ) control . Strains were synchronized by hypochlorite treatment followed by allowing C . elegans to grow for approximately 2 . 5 days at 20°C . All experiments were carried out with young adult hermaphrodites at ~ 23°C , unless otherwise mentioned . A complete list of strains utilized in this study is given in Tables C and D in S1 Text . OP50 Escherichia coli was obtained from the C . elegans Genetics Center ( University of Minnesota , Minneapolis , MN , USA ) . Tables A and B in S1 Text lists all the plasmids and constructs used in this study , Table E in S1 Text lists the primers used to perform genotyping and Table F in S1 Text lists the primers used to make the different transgenes used in this study . All the plasmids were generated using standard restriction digestion based cloning strategy and sequenced before use in experiments . Previously described microinjection techniques were used to generate stable transgenic C . elegans lines carrying extra- chromosomal DNA arrays using either pmyo-3::mCherry , pmyo-2::GFP or popt-3::mCherry as co-injection markers [90] . The three casy-1 isoforms studied in the manuscript are mentioned on the Wormbase based on EST evidence . The sequence information has been obtained from Wormbase to design isoform-specific primers to amplify the promoter sequences ( ~ 3 kb upstream of the translation start codon ) as well as the coding region . All the assays were performed with the experimenter blind to the genotypes . Each assay was performed at least three times as indicated at the base of each bar with > than 20 animals for each replicate . Animals were immobilized with 30mg/ml 2 , 3-butanedione monoxamine ( Sigma ) on 2% agarose pads . All quantitative imaging was done using Zeiss AxioImager microscope with a 40x or 63x 1 . 4 NA Plan APOCHROMAT objective equipped with a Zeiss AxioCam MRm CCD camera controlled by Axiovision software ( Zeiss Micro-imaging ) . For comparing WT animals with casy-1 mutants , >25 C . elegans were analyzed for each genotype [38] . For the morphological analysis of GABAergic neurons ( juIs76 [Punc-25::GFP] ) and cholinergic neurons ( nuIs321 [Punc-17::mCherry] ) , z-stacks of the entire C . elegans ( from overlapping fields of view ) were taken using a 63x objective and a 10 μm optical slice ( 20 slices at 0 . 5 micron distance ) . Image J software was used to derive maximum intensity projection images from z-stacks . These images were then analyzed for gross morphological analysis of neurons . All fluorescent SV marker imaging was done with the 63x objective . Animals were imaged for the DNC in the posterior portion of the C . elegans halfway between the vulva and the tail . For fluorescent analysis , image stacks were taken ( approximately 10 μm ) and the maximum intensity projections were obtained using Image J software . For the analysis of fluorescence intensity a freehand line was drawn ( approximately 100 μm ) along the DNC and the intensity values were obtained for each animal indicated . These values were normalized to WT levels and then an average plot has been drawn to determine the statistical difference in the intensity . For all the quantitative analysis , identical camera gain , exposure settings , and fluorescence filters were used for a particular transgenic line . For all the Figures , an average of the values for each C . elegans in the data set ± S . E . M . is plotted . Statistical difference between WT and mutant values was determined using the Student's t-test ( p ≤ 0 . 05 ) in Graph Pad Prism 7 . Graphs of punctal and axonal fluorescence show data normalized to WT values . Transcriptional reporter expression , superecliptic pHluorin , BiFC and co-localization imaging was done using Leica HC PL APO 63x/ TCS SP8 confocal microscope ( Leica Microsystems ) with Multi-Ar ( 457 , 488 , and 515 nm ) , and He-Ne ( 543 and 633 nm ) laser lines and HyD detectors . For co-localization analysis , image stacks were taken ( four times of average for 1024x1024 scan format at speed 400 Hz ) and maximum intensity projections were obtained for each channel separately . Maximum intensity projections were then merged to obtain co-localization images using the Image J software . Strains for electrophysiology were maintained on plates seeded with HB101 Escherichia coli at 20°C . Adult C . elegans were immobilized on Sylgard coated coverslips with cyanoacrylate glue . Dissections and whole-cell recordings were performed as described previously [91–93] . Statistical significance was determined using the one-way ANOVA followed by Dunnett’s test for comparison of mean frequency and amplitude for endogenous EPSCs and IPSCs . Optogenetic experiments were performed as described previously [94] . All- trans retinal ( ATR ) plates were prepared fresh and used within one week . ATR plates were made by spotting NGM plates with 50 μl of OP50 E . coli containing 0 . 8mM ATR and allowed to grow overnight at 37° C in dark . Control plates were made by seeding NGM plates with 50μl of OP50 lacking ATR . Transgenic animals carrying GABAergic Channelrhodopsin zxIs3 [Punc-47:: ChR2 ( H134R ) ::YFP + lin-15 ( + ) ] were synchronized and grown on ± ATR plates at 20°C . For the assay , ~ 10 young adult hermaphrodites were picked on fresh seeded NGM plates . For all analysis , a 20- second video was made using the Zeiss Lumar V12 fluorescence Stereomicroscope . Video recording was started as soon as the animals were seen crawling , blue light was turned on within 4 seconds and then left on for the duration of recording . The light intensity was approximately 57 . 5 mW/cm2 from HXP 120V light source . Two frames were selected from all the videos for each animal , one before blue light was turned on and one after that with the maximum relaxation of the C . elegans . Analysis was performed as described previously [54] . In brief , Image J on a Wacom Bamboo tablet and stylus were used to trace a freehand line from the nose tip down to the posterior-most point of the C . elegans . The length of the animal before and after exposure to blue light was measured and difference in these lengths divided by the starting C . elegans length was determined . This data was then plotted as percentage change in body length ( relaxation ) . FRAP was performed on transgenic WT and casy-1 ( tm718 ) mutants carrying GABAergic nuIs376 [Punc-25:: SNB-1::GFP] or cholinergic nuIs152 [Punc-129::SNB-1::GFP] . For RNAi- based FRAP experiments , the F2 generation of C . elegans exposed to unc-104 RNAi were used . FRAP experiments were carried out using Leica TCS SP8 confocal microscope . For the FRAP experiments , C . elegans were immobilized on 10% agarose pads containing 0 . 2μl of 0 . 1μm polystyrene microspheres ( Polysciences ) . To account for the focal drift , image stacks were taken , and maximum intensity projections were obtained using Image J . For this experiment , three frames were taken for one representative puncta in the posterior half of the animal , followed by 100 iterations of photobleaching in the defined region ( 25% power of a 488 nm Argon laser with bleaching power of 45% ) . This was followed by monitoring 10 iterations of recovery every 25 seconds for up to six minutes . The intensities were normalized against a non-bleached ROI within the same animal . RNAi experiments were carried out as described previously [95] . All assays were performed in eri-1 ( mg366 ) ; lin-15B ( n744 ) background which makes animals more sensitive to RNAi in the nervous system[96] . Briefly , animals were raised on bacteria expressing double-stranded RNA containing casy-1 , unc-104 or empty vector for two generations . F2 generation animals were analyzed on acute Aldicarb assays in duplicates . For GST- pull down experiment soluble C-terminal of casy-1 cDNA ( amino acids 48–160 ) and casy-1 ( Δkif ) cDNA was subcloned into the pGEX-KG vector and expressed as N-terminal GST fusion proteins in Escherichia coli BL21 cells . Cells were grown at 37°C to an OD600 nm of 0 . 6 , induced with 0 . 1 mM isopropylthiogalactoside ( IPTG ) and grown further for 16 hours at 18°C . Harvested cells were resuspended in a lysis buffer containing 50 mM Tris at pH 7 . 5 , 300 mM NaCl , Protease inhibitor ( Roche ) and DNase I ( Sigma ) and disrupted by freeze–thawing followed by short sonication . The cell lysate was centrifuged at 14 , 000 rpm for 20 minutes at 4°C and the cleared supernatant was incubated with glutathione beads ( Sigma ) for 2 hours at 4°C . The beads were washed with lysis buffer three times . The beads were then packed into a syringe column . GST-tagged protein immobilized on glutathione–Sepharose beads was incubated with the supernatant from E . coli BL21 cells expressing UNC-104-HA for 2 hours at 4°C . The supernatant was prepared as mentioned above . The beads were then washed extensively five times . After extensive washing with lysis buffer , bound proteins were added directly into 5×SDS/PAGE sample buffer , boiled for 5 min , separated by SDS/PAGE and subjected to immunoblot analysis . Western blot using Anti-HA Antibody ( 1∶1000 ) was then performed to identify interaction between CASY-1 and UNC-104 . For BiFC studies , transgenic strains containing following BiFC pairs were generated ( 1 ) Punc-104::UNC-104::VN173 and Punc-104::UNC-104::VC155 ( positive control ) ( 2 ) Punc-104::UNC-104::VN173 and Empty VC155 ( negative control ) ( 3 ) Punc-104::UNC-104::VN173 and Punc-25::CASY-1C ( amino acids 1–160 ) ::VC155 and ( 4 ) Punc-104::UNC-104::VN173 and Punc-25::CASY-1C ( ΔKIF ) ::VC155 . The transgenic lines were imaged using Leica HC PL APO 63x/ TCS SP8 confocal microscope in the posterior portion of the C . elegans halfway between the vulva and the tail . The transgenic lines obtained showed a discontinuous expression along the entire VNC probably due to low efficiency of extrachromosomal arrays . For in vivo time-lapse imaging of the motor neuron commissures , 1-day adult hermaphrodites with the GABAergic nuIs376 [Punc-25::SNB-1::GFP] or cholinergic nuIs152 [Punc-129::SNB-1::GFP] transgene were immobilized in 3 mM tetramisole in M9 and mounted on a 5% agarose pad . Imaging was performed in the posterior commissures . SNB-1::GFP time-lapse images were obtained with an Olympus IX83 microscope ( Olympus , Tokyo , Japan ) using a Plan Apochromat objective ( 100X , 1 . 4 NA ) attached with a spinning disk confocal head ( CSU22; Yokogawa , Tokyo , Japan ) and equipped with an electron-multiplying charge-coupled device ( EMCCD ) camera ( ImagEM X2 EM-CCD , Hamamatsu ) . A 488 nm laser ( 25 mW ) was used at 10% power for imaging ( 100X objective , 300 ms exposure time ) . Moving particles were defined as puncta that were displaced by more than 3 pixels in less than 5 consecutive time frames . The flux of particles was calculated as the total number of puncta moving in either direction in an entire kymograph , then normalized to a 10 μm region , further normalized to time ( 1 minute ) . The Unit of flux here is number of events/10μm/min . Movies were acquired at a constant frame rate of 3 frames/s for a total of 700 ( again average number- varies across movies ) frames . Each frame was 512x512 pixels dimension . Kymographs were generated and analyzed using ImageJ software , version 1 . 41 ( National Institutes of Health , Bethesda , MD ) , and statistical significance was determined using a two-way ANOVA with Bonferroni's multiple comparison post-test . All statistical analysis were performed using GraphPad Prism V7 . Experimental data are shown as mean ± S . E . M . Statistical comparisons were done using the Student's t-test , two-way ANOVA or one-way ANOVA with Bonferroni's multiple comparison or Dunnett’s post-test . A level of p<0 . 05 was considered significant .
|
GABA acts as a major inhibitory neurotransmitter in both vertebrate and invertebrate nervous systems . Despite the potential deregulation of GABA signaling in several neurological disorders , our understanding of the genetic factors that regulate GABAergic synaptic transmission has just started to evolve . Here , we identify a role for a cell adhesion molecule , CASY-1 , in regulating GABA signaling at the C . elegans NMJ . We show that the mutants in casy-1 have reduced number of GABA vesicles at the synapse resulting in less GABA release from the presynaptic GABAergic motor neurons . Further , we show that the shorter isoforms of the casy-1 gene; casy-1b and casy-1c that carry a potential kinesin-motor binding domain are responsible for maintaining GABAergic signaling at the synapse . We show a novel interaction of the CASY-1 isoforms with the C- terminal of the UNC-104/KIF1A motor protein that mediates the trafficking of GABAergic synaptic vesicle precursors to the synapse , thus maintaining normal inhibitory signaling at the NMJ .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"invertebrates",
"fluorescence",
"imaging",
"medicine",
"and",
"health",
"sciences",
"neurochemistry",
"caenorhabditis",
"nervous",
"system",
"immunology",
"electrophysiology",
"neuroscience",
"animals",
"motor",
"neurons",
"animal",
"models",
"caenorhabditis",
"elegans",
"model",
"organisms",
"clinical",
"medicine",
"experimental",
"organism",
"systems",
"neurotransmitters",
"hypersensitivity",
"research",
"and",
"analysis",
"methods",
"cholinergics",
"imaging",
"techniques",
"animal",
"cells",
"gamma-aminobutyric",
"acid",
"biochemistry",
"cellular",
"neuroscience",
"eukaryota",
"cell",
"biology",
"anatomy",
"synapses",
"clinical",
"immunology",
"neurotransmission",
"physiology",
"neurons",
"nematoda",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"neurophysiology",
"organisms"
] |
2018
|
The C-terminal of CASY-1/Calsyntenin regulates GABAergic synaptic transmission at the Caenorhabditis elegans neuromuscular junction
|
Rapid place encoding by hippocampal neurons , as reflected by place-related firing , has been intensely studied , whereas the substrates that translate hippocampal place codes into behavior have received little attention . A key point relevant to this translation is that hippocampal organization is characterized by functional–anatomical gradients along the septotemporal axis: Whereas the ability of hippocampal neurons to encode accurate place information declines from the septal to temporal end , hippocampal connectivity to prefrontal and subcortical sites that might relate such place information to behavioral-control processes shows an opposite gradient . We examined in rats the impact of selective lesions to relevant parts of the hippocampus on behavioral tests requiring place learning ( watermaze procedures ) and on in vivo electrophysiological models of hippocampal encoding ( long-term potentiation [LTP] , place cells ) . We found that the intermediate hippocampus is necessary and largely sufficient for behavioral performance based on rapid place learning . In contrast , a residual septal pole of the hippocampus , although displaying intact electrophysiological indices of rapid information encoding ( LTP , precise place-related firing , and rapid remapping ) , failed to sustain watermaze performance based on rapid place learning . These data highlight the important distinction between hippocampal encoding and the behavioral performance based on such encoding , and suggest that the intermediate hippocampus , where substrates of rapid accurate place encoding converge with links to behavioral control , is critical to translate rapid ( one-trial ) place learning into navigational performance .
A classical distinction in animal learning theory is that between “learning” and “performance” [1 , 2] . The processes involved in encoding and storing new information are conceptually distinct from those involved in translating that information into useful behavior . The recent preoccupation with plasticity and encoding mechanisms has sometimes led to this distinction being forgotten , but there are dangers in doing so [3] . The present study focuses on the neuroanatomical substrates of the learning–behavior translation with particular attention to the issue of how rapid place learning results in effective navigational behavior . Different theories hold that the hippocampus mediates certain forms of rapid learning , including place learning [4–11] . For example , when rats explore a novel environment , hippocampal principal neurons rapidly form place codes , as reflected by place-specific firing [9 , 12–14] . Recent research has focused on how neocortical visuospatial inputs , entering the hippocampus from the entorhinal cortex [15–18] , are processed by different subregions along the transverse and longitudinal axes of the hippocampus to mediate place representations . This work has revealed that different subregions along the hippocampal transverse axis ( dentate gyrus , CA3 , and CA1 ) make distinct computational contributions , including rapid encoding and pattern completion ( CA3 ) [19–22] , pattern separation ( dentate gyrus and CA3 ) [23 , 24] , and the comparison of new and stored information ( CA1 ) [25–27] . In addition , there is a gradient in the precision of visuospatial encoding along the hippocampal longitudinal , or septotemporal , axis that runs from the septal pole , close to the septum , to the temporal pole , close to the amygdala . Main hippocampal connections display septotemporal topographical gradients; that is , they gradually get weaker toward the septal or temporal pole , so that they are mainly restricted to approximately one- to two-thirds starting from either pole , and thereby an anatomical differentiation emerges into three partly overlapping domains with different sets of connectivity: a septal and temporal region , and , between them , an intermediate region ( Figure 1A ) [28–32] . With respect to the precision of visuospatial encoding , it is critical that the septal to intermediate hippocampus exhibit strong connectivity to the dorsolateral domain of the entorhinal cortex , which receives strong visuospatial neocortical inputs and where neurons , so-called grid cells , represent visuospatial information at a fine scale ( grid-like arranged firing fields of down to 20-cm diameter whose centers are spaced at as little as 30 cm apart ) ; in contrast , the temporal hippocampus is mainly connected to the ventromedial domain of the entorhinal cortex , which receives little visuospatial neocortical input and where fine-grained grid-cell firing patterns are replaced by coarser ones ( firing fields of up to 3-m diameter spaced at 5 m ) ( Figure 1A , top left ) [15 , 18 , 33 , 34] . Consistent with this , precise place-field codes for small areas ( ca . 10- to 20-cm diameter ) in conventionally sized recording environments ( ca . 1 m × 1 m or smaller ) are restricted to the septal to intermediate hippocampus [25 , 35–40] . In contrast , most neurons in the temporal hippocampus do not display precise place codes; they show larger and less accurate place fields ( several meters in diameter ) when recordings are performed in relatively large environments ( e . g . , an 18-m linear track ) [38] . This important work has focused on hippocampal place encoding . However , how are hippocampal place codes translated into behavior ? This translation may involve different anatomical routes as a function of whether information has been recently and rapidly acquired or been consolidated in the neocortex through incremental learning and/or over time [10 , 41 , 42] . With respect to rapidly and recently acquired place memory , particularly in circumstances when relevant information is changing frequently and cannot be consolidated into neocortex , the translation into behavior is likely to involve hippocampal links to medial prefrontal cortical areas , such as the prelimbic and infralimbic cortex , and subcortical sites , such as the mediodorsal striatum , the nucleus accumbens , the amygdala , lateral septum , and hypothalamus . This is because these brain regions provide access to brain-stem sites mediating motor responses [43–46] and are considered to play key roles in behavioral-control processes ( including emotional , motivational , sensorimotor , and executive functions ) [30 , 31 , 45 , 47–54] . A key idea is that whereas the septal to intermediate hippocampus mediate fine-grained , accurate place encoding ( through interactions with the dorsolateral domain of the entorhinal cortex ) , strong neuroanatomical and functional links to behavioral control ( through connections to prefrontal cortex and subcortical sites ) are mainly provided by intermediate to temporal hippocampal regions [29–32 , 45 , 55–61] ( for neuroanatomical details , see Figure 1A and accompanying legend ) . It follows that the intermediate hippocampus , where substrates of accurate place encoding converge with direct links to behavioral control , may be critical for the translation of rapid place learning into behavior [47] . Based on these considerations , we hypothesized that the intermediate hippocampus would be both necessary and largely sufficient for behavior on tasks requiring rapid place learning . In contrast , neither the septal nor temporal pole of the hippocampus , each comprising only one of the two complementary sets of functional connectivity ( i . e . , to dorsolateral domain of entorhinal cortex or to prefrontal cortex and subcortical sites ) , would sustain such performance . Nevertheless , the septal pole , through its connectivity with the dorsolateral domain of the entorhinal cortex , should be able to mediate the rapid encoding of accurate place representations ( even though unable , on its own , to translate this information into action ) . Importantly , this latter prediction may appear paradoxical from theoretical viewpoints that focus on hippocampal encoding mechanisms alone , but it follows directly from the perspective put forward here emphasizing the distinction between hippocampal encoding and behavioral performance based on such encoding . To test these hypotheses , we examined in rats the impact of highly specific partial hippocampal lesions , sparing distinct parts along the septotemporal axis , on performance of a watermaze task requiring rapid , one-trial , learning of a new place every day and on electrophysiological models of rapid encoding in the septal hippocampus , long-term potentiation ( LTP ) [62] , and place-related cell firing . An important qualification that influenced our experimental plan is that rapid place learning may have distinct mechanisms from incremental learning over many trials [21–23 , 41 , 42 , 63–67] . Indeed , in contrast to our predictions for a one-trial learning paradigm with a daily changing goal location , previous watermaze studies using reference memory paradigms , in which the same location is learned over many trials across several days , have found that small residuals of hippocampal tissue , especially at the septal pole , can be sufficient to sustain good performance [39 , 57 , 68 , 69] . In fact , slow learning can even occur in the absence of the hippocampus [66 , 67] . Through incremental training , accurate place information might be acquired [42] and/or consolidated [41] in the neocortex and , from there , be translated into behavior [46] . This route would not require direct hippocampal links to behavioral control . Therefore , one of our watermaze experiments contrasted the impact of hippocampal lesions on performance based on rapid , one-trial , learning with that based on incremental learning .
At first glance , our results may appear inconsistent with previous findings that rats with only 20%–40% of residual hippocampal volume at the septal or temporal pole can express similar place memory in the watermaze as control groups [68 , 69] . However , in these studies , incremental learning was possible , as the rats were trained to the same place over many trials . To compare performance based on rapid versus incremental learning , we therefore trained the different lesion groups with one constant platform location for 5 d , with eight trials per day , and trials 2 and 6 of each day run as probes to closely monitor performance . On day 6 , one additional probe trial was run to assess incrementally acquired long-term memory independent of within-day learning . On day 7 , a subset of the rats ( the last four of five replications , a total of 72 rats: 20 sham-operated , 13 with residual intermediate hippocampus , 15 with residual temporal pole , 10 with residual septal pole , and 14 with complete hippocampal lesions ) was retested for performance based on rapid place learning , with a novel platform location and four trials ( trial 2 run as probe ) ; this retest was included to rule out that nonspecific recovery due to extended postsurgical training might account for any better performance after incremental place learning ( Figure 3A ) . The differential effects of hippocampal damage on performance based on rapid versus incremental place learning were strikingly revealed by search preference on trial 2 of the first day , i . e . , after one learning trial , and on trial 1 of day 6 , i . e . , after 40 trials run across the preceding 5 d ( Figure 3B ) . After only one learning trial , rats with sham lesions or residual intermediate hippocampus focused their search on the correct zone , whereas rats with only the septal or temporal pole , or with complete hippocampal lesions were impaired ( Figure 3B , left ) , replicating the results from experiment 1 ( compare Figure 2D ) ; in contrast , after slow , incremental learning over 40 trials , groups did not differ anymore ( Figure 3B , right ) ( interaction group × probe trial [day 1 , trial 2 , vs . day 6 , trial 1]: F4 , 84 = 2 . 5 , p < 0 . 05; main effect of group , day 1 , trial 2: F4 , 84 = 11 . 0 , p < 0 . 0001 , post hoc comparisons: see figure; main effect of group , day 6 , trial 1: F4 , 84 = 1 . 4 , p > 0 . 22 ) . Interestingly , the sham-operated rats and all groups with partial hippocampal lesions showed some rapid within-day performance improvements that were not carried over to the next day , so that latencies on trial 1 of a day tended to be higher than on the last trial of the preceding day and percentage of time in the correct zone tended to be lower on the first probe of a day ( trial 2 ) than on the last probe ( trial 6 ) of the previous day . In contrast , rats with complete hippocampal lesions gradually improved across trials and days ( Figure S3 ) . In previous studies , using incremental training in the watermaze , rats with residual 20%–40% of hippocampal volume at the septal pole did not differ from sham-lesioned rats at a stage when performance was still hippocampus dependent , i . e . , impaired in rats with complete hippocampal lesions [69] ( under certain training conditions , this also pertained to rats with 20%–40% of hippocampal volume restricted to the temporal pole [68] ) . These findings were confirmed in the present experiment: on trial 1 of day 4 , paths were longer in rats with complete hippocampal lesions ( F4 , 84 = 2 . 76 , p < 0 . 03 ) , as compared to sham-operated rats ( p < 0 . 01 ) and the partial-lesion groups ( p < 0 . 03 ) , excepting rats with remnants at the temporal pole for which the difference just failed to reach significance ( p = 0 . 054 ) ; the partial lesion groups did not differ from each other or the sham-lesioned rats ( p > 0 . 3 ) ( Figure S3A ) . These results demonstrate that hippocampal contributions are especially required for behavior requiring rapid place learning; if incremental place learning is possible , good navigational performance can be achieved with hippocampal residuals at the septal or temporal pole , and eventually even without a hippocampus . In stark contrast to the similarly good performance across groups after incremental place learning ( Figure 3B , right ) , there were again marked group differences when rats were subsequently retested for performance based on rapid place learning on day 7 ( Figure 3C; for path-length data , see Figure S4 ) . During the probe on trial 2 , after one training trial to a novel location , only the sham-operated rats and those with a residual intermediate hippocampus significantly preferred the correct zone; rats with only the septal or temporal pole of the hippocampus did not differ from chance , and rats with complete hippocampal lesions spent even less time in the correct zone than expected by chance ( as they kept searching for the platform in the previous , incrementally learnt location , but had not learnt the novel location ) ( main effect of groups: F4 , 67 = 18 . 52 , p < 0 . 0001; between-groups post hoc comparisons and comparisons to chance: see Figure 3C ) . These results corroborate the special importance of the hippocampus for performance based on rapid place learning . Interestingly , even though the group with an intact intermediate hippocampus was still the best of all groups that had any damage to the hippocampus , it was significantly impaired as compared to the sham-operated group ( p < 0 . 0025 ) , possibly related to interference from the preceding incremental place-learning task . Rats with hippocampal residuals at the septal pole showed poor performance on the rapid place-learning task , whereas rats with a residual intermediate hippocampus displayed largely intact performance . According to our hypothesis , this finding does not reflect that a residual septal pole is less capable of accurate place encoding/learning than a residual intermediate hippocampus ( but rather that a residual septal pole , lacking the links of the intermediate hippocampus to behavioral-control sites , is incapable of translating place learning into performance ) . Indeed , in the intact hippocampus , neurons at the septal pole encode visuospatial information at even higher precision than neurons in the intermediate region [37 , 38 , 40] . We predicted that such encoding in the septal hippocampus would be unaffected by neurotoxic lesions to the rest of the hippocampus , given that such lesions should not affect the relevant entorhinal–hippocampal interactions [16 , 18 , 39 , 71] . To test this point , we examined properties of the residual hippocampal circuitry at the septal pole using electrophysiological models of information encoding , predicting that synaptic plasticity and place-field encoding would be essentially normal . In experiment 3 , evoked field potentials in the perforant path–dentate gyrus pathway , a main entorhinal input to the hippocampus , were recorded from the septal hippocampus of anesthetized rats that were sham-lesioned , i . e . , had an intact hippocampus ( mean ± SEM: 100 ± 3 . 5% , range: 84 . 4%–107 . 5%; n = 6 ) , or had received partial hippocampal lesions sparing only the septal pole ( 34 . 1 ± 1 . 7% of control hippocampus , 27 . 1%–41 . 7%; n = 7 ) ( Figure 4A ) . Input–output curves indicated no significant difference in field-potential slopes between these groups ( F1 , 11 = 2 . 72 , p > 0 . 12 ) ( Figure 4B , left ) . However , in lesioned rats , perforant-path stimulation triggered neuronal firing more readily than in sham-operated rats , as indicated by significantly higher population spike amplitude and population spike/slope ratio ( main effects of group: F1 , 11 > 7 . 16 , p < 0 . 02 ) across all stimulation intensities , except for the lowest ( interactions group × intensity: F9 , 99 > 2 . 02 , p < 0 . 05 ) ( Figure 4B , middle and right ) . These findings may reflect that lesion of the temporal and intermediate hippocampus removed feed-forward inhibition by longitudinally projecting inhibitory interneurons , which have been suggested to support coordination of neuronal firing along the septotemporal axis in the intact hippocampus [72–74] . Feedback inhibition , as indicated by paired-pulse inhibition [75] , could readily be demonstrated in hippocampal residuals at the septal pole ( Figure 4C ) , ruling out a general hyperexcitability . Importantly , tetanization resulted in robust LTP in both groups ( Figure 4D ) . Slopes measured 55–60 min after the tetanus were significantly increased as compared to the 5 min preceding the tetanus in both groups ( t > 2 . 87 , p < 0 . 05 ) and did not differ between groups ( t11 = 1 . 15 , p = 0 . 27 ) . Thus , hippocampal residuals at the septal pole displayed normal synaptic plasticity . In experiment 4 , single-unit recordings revealed accurate and stable place-related firing in hippocampal residuals at the septal pole ( Figure 5A ) in rats that were foraging in large open fields surrounded by prominent distal visual cues . Starting with a highly familiar environment , 42 putative pyramidal neurons were recorded from three rats with partial hippocampal lesions sparing only the septal pole ( mean ± SEM: 30 . 4 ± 2 . 3% of control hippocampus , range: 25 . 8%–33 . 02% ) . These neurons exhibited highly place-related firing that was stable between two recording trials in the familiar environment ( typically separated by 30 to 60 min ) ( Figures 5B and S5A ) ; spatial firing was similar to that of pyramidal neurons ( n = 28 ) recorded from the septal pole of an intact hippocampus ( 100 ± 11 . 5% , 79 . 9%–119 . 8% ) in control rats ( n = 4 ) ( Figure S5B ) . Quantitative measures of place-related firing ( average and peak rates , spatial information , and sparsity ) and its stability between two successive trials in the familiar environment ( correlation , changes in average and peak firing rate , movement of firing peak ) were calculated ( mean ± SEM ) and compared between groups ( data of different cells recorded from the same rat were averaged , so that each rat contributed one value to the group means ) ( Table S1 ) . Only the stability measures differed , with a lower between-trials correlation of firing patterns in cells from septal remnants ( 0 . 75 ± 0 . 03 , as compared to 0 . 88 ± 0 . 02 in the intact hippocampus; t5 = 3 . 58 , p < 0 . 02; correlation coefficients were subjected to Fisher's z′-transformation before the t-test ) . This difference could be accounted for by a relatively high between-trials change in peak firing rate in place cells of septal remnants ( 0 . 28 ± 0 . 04; compare also Figures 5A and S5A ) , which was twice as high as in cells from an intact hippocampus ( 0 . 14 ± 0 . 03; t5 = 3 . 17 , p < 0 . 025 ) . Importantly , the position of peak firing was equally stable in both groups ( movement of firing peak between trials in centimeters: septal remnants , 14 . 3 ± 3 . 6; intact hippocampus , 16 . 5 ± 4 . 2 , t5 < 1 ) . Finally , when exposed to a new environment , CA1 pyramidal neurons ( n = 9 ) in a hippocampal residual at the septal pole rapidly changed their firing patterns reflecting normal remapping [13] ( Figures 5C and S6 ) . Thus , overall , neurons in hippocampal residuals at the septal pole displayed normal accurate and rapid place encoding . Experiment 5 addressed whether the convergence of substrates for rapid place encoding with those for behavioral control in the intermediate hippocampus was really critical for performance based on rapid place learning , as implied by our conceptual framework . Alternatively , independent , parallel contributions of substrates mediating rapid place encoding and of links to behavioral control at the septal and temporal tip of the hippocampus , respectively , could mediate performance . To decide between these two alternatives , our last experiment tested whether residuals at the septal and temporal tip ( ca . 20% of total hippocampal volume spared at each tip ) , separated by a lesion to the intermediate hippocampus , could sustain performance . Thus , experiment 5 is the “mirror image” of experiment 1: whereas experiment 1 established that hippocampal tissue restricted to the intermediate hippocampus was sufficient for performance based on rapid place learning , experiment 5 tested whether damage restricted to this area causes an impairment in performance . Importantly , whereas different septotemporal levels are connected through intrahippocampal longitudinal projections , there is no evidence for direct links between the septal and temporal 20% of the hippocampal volume , and the only intrahippocampal connection between these septal and temporal tips may be by way of the intermediate region , which receives afferents from the septal tip and projects to the temporal tip [28 , 76–80] . Thus , the septal and temporal tips spared by the lesions made to the intermediate hippocampus in the present experiment would essentially be disconnected . Rats were pretrained on the rapid place-learning task as described for experiment 1 and then divided into two performance-matched groups ( Figure S7 ) . One group received sham surgery ( hippocampal volume , mean ± SEM: 100 . 0 ± 2 . 1% , range: 90 . 7%–115 . 0%; n = 12 ) ; the other one had lesions to the intermediate hippocampus , sparing approximately 20% of hippocampal volume each at the septal ( 19 . 9 ± 1 . 2% , 12 . 3%–25 . 4% ) and the temporal ( 22 . 1 ± 0 . 75% , 8 . 3%–27 . 1% ) tip of the hippocampus ( n = 11 ) ( Figure 6A and 6B; Video S5 ) ( n refers to the rats included in data analysis; for additional lesion analysis , see Text S1 , Supplementary Results 1 ) . Lesions to the intermediate hippocampus virtually abolished performance ( Figure 6C ) . Performance did not significantly depend on the retention delay between trial 1 and 2 ( Figure S8 ) , and group differences were delay-independent ( interactions involving groups × delays: F < 1 ) , so the analysis focused on data averaged over both delays . In contrast to sham-operated rats , rats without the intermediate hippocampus showed no search preference for the correct zone on probe trials ( main effect of group: F1 , 21 = 10 . 9 , p < 0 . 004 ) and no savings between trial 1 and 2 ( interaction groups × trials: F3 , 63 = 6 . 8 , p < 0 . 001; main effect of group on savings: F1 , 21 = 14 . 3 , p < 0 . 002; comparisons to chance: see figure ) ( Figure 6C ) . Thus , separated hippocampal residuals at the temporal and septal tip cannot sustain performance . In each of the watermaze experiments ( experiments 1 , 2 , and 5 ) , lesions affecting the temporal half of the hippocampus increased swim speed ( Text S1 , Supplementary Results 2 ) . This increase is in line with previous observations [56] and corroborates the close association of the temporal half of the hippocampus with sites mediating motor function [55] .
Hippocampal lesions caused a quite different pattern of results on the rapid , one-trial , place-learning task , in which the goal location changed every day , than on the incremental-learning task , in which rats were trained to the same location over many trials across days and which belongs to the more commonly used reference-memory paradigms . A conceptually important difference between these two watermaze tasks relates to the relevance of hippocampal information encoding and neocortical memory acquisition/consolidation . On the rapid-learning task , the goal location is constantly changing from day to day , requiring the rapid encoding of stimuli and their relations , for which the hippocampus and its synaptic plasticity are critical [8 , 9 , 21–23 , 41 , 42 , 64 , 66 , 70 , 81 , 82] . Information about the goal location must be constantly updated by hippocampal encoding mechanisms and , from the hippocampus , secure direct access to behavioral control systems . Consistent with this , we found that when rapid , one-trial , learning of a novel place is required , the hippocampus , especially the intermediate part , is essential for effective performance ( independent of retention delay , see Text S1 , Supplementary Discussion 1 ) . On the incremental-learning task , in contrast , all relevant information is stable over time . It has been suggested that under these circumstances , information can be consolidated from the hippocampus into the neocortex [41 , 83 , 84] or gradually be acquired by the neocortex [42] . From neocortical storage sites , effective behavioral control is possible via connections bypassing hippocampal pathways [46] . Although hippocampal circuitry may normally contribute to incremental place-learning tasks , and even small residuals of such circuitry , especially at the septal pole , may speed up acquisition ( discussed below ) , our experiment 2 demonstrated that even rats with complete hippocampal lesions could eventually , albeit at a much slower rate than a control group , come to display accurate place navigation ( as shown on probe trials by their focused search preference in a small 40-cm-diameter zone centered on the platform location ) . This confirms and extends previous demonstrations of relatively intact performance in the watermaze after extended incremental training in rats with complete hippocampal lesions [66 , 67] ( also see Text S1 , Supplementary Discussion 2 ) . Humans with hippocampal lesions , who show marked impairments in place and declarative memory , can also come to express accurate place [85] and semantic [86 , 87] memory , when incremental learning is possible . Without the hippocampus , initially coarse neocortical representations of relevant information , for example of places in the entorhinal cortex [15 , 17] , may slowly be sharpened into accurate , nonoverlapping representations through incremental learning [42] and then be translated into behavior via direct neocortical connections . Hippocampal circuitry may normally contribute to incremental place-learning tasks; however , it is not absolutely required , and relatively good performance on such tasks can eventually be achieved without the hippocampus , albeit very slowly ( also compare [66] ) . Rats with hippocampal tissue restricted to the septal or temporal pole were markedly impaired on all tests requiring rapid , especially one-trial , place learning . Both groups exhibited deficits in the main performance measure , search preference during probes , whereas only the rats with tissue restricted to the septal pole showed impaired path-length savings . A number of previous studies have found that rats with partial hippocampal lesions sparing only the septal pole can display efficient performance on place-learning tasks , whereas rats with residuals at the temporal pole cannot ( for review , see [39 , 57] ) . For example , after 32 trials ( eight trials/day , 4 d ) to the same platform location in the watermaze , rats with hippocampal remnants of 20%–40% of total volume at the septal pole showed relatively normal performance , whereas rats with even 40%–60% spared hippocampal volume at the temporal pole were substantially impaired [68 , 69] . What accounts for this advantage of rats with septal hippocampal sparing on the reference-memory version of the watermaze task ? Residual septal hippocampal circuitry is capable of rapid place encoding , as revealed by our place-cell recording experiments , even though it lacks the connectivity to relate place codes directly to behavioral-control functions . However , as all relevant place information on the reference-memory version of the watermaze task is stable , place information acquired within hippocampus may , through incremental learning and the process of systems-level consolidation , become established , or “interleaved , ” into neocortical networks [8 , 41 , 83 , 84 , 88 , 89] , which can then translate such stored information into behavior . A residual septal pole may accelerate performance acquisition on watermaze reference-memory tasks , because neocortical learning aided by rapid septal hippocampal place encoding is faster than purely neocortical learning ( even though eventually even rats with complete hippocampal lesions can come to show good performance ) . Importantly , however , several studies have found that rats with 30%–50% of hippocampal volume at the septal pole show performance deficits at early stages of incremental place learning tasks [90] , i . e . , when performance improvement likely depends on rapid learning and its translation into action through direct hippocampal links to behavioral control , and that these deficits diminish with additional training [68 , 91 , 92]; performance deficits have also been indicated on watermaze tests of one-trial place learning as reduced savings [92 , 93] ( but see [56] ) . Although the boundary conditions under which the septal or temporal pole of the hippocampus can sustain performance on place-learning tasks remain to be clarified ( also see [68] ) , our new findings demonstrate that neither region alone can sustain normal navigational performance based on rapid , one-trial , place learning . Our findings reflect , we suggest , that the temporal region of the hippocampus cannot form accurate place codes , as it lacks close links to the dorsolateral domain of the entorhinal cortex , whereas the septal region , due to its connectivity with this part of the entorhinal cortex , can rapidly form accurate place codes , but not translate them into behavioral control , due to the lack of connectivity to prefrontal cortex and subcortical sites ( see Introduction and Figure 1A ) . Several lines of evidence , comprising recordings of single-unit firing and of multisite coordinated activity , as well as studies using lesion and pharmacological manipulations of relevant brain sites ( including disconnection approaches ) , support the notion that hippocampal interactions with the dorsolateral domain of the entorhinal cortex [16 , 94–96] and with medial prefrontal cortex and subcortical sites , specifically the mediodorsal striatum and nucleus accumbens [47 , 48 , 51 , 97–104] , are important for performance on place-learning tasks . In support of our emphasis on the importance of hippocampal–prefrontal/subcortical interactions in interpreting our results , the following findings are noteworthy . First , lesions to medial prefrontal cortex or mediodorsal striatum in rats [50] or glutamate-receptor blockade in the nucleus accumbens of mice [105] impair performance on watermaze rapid place-learning tasks similar to the one used in the present study . Second , using crossed unilateral lesions or pharmacological manipulations of the relevant sites , it was demonstrated that disconnection of the hippocampus from medial prefrontal cortex [48 , 103 , 104] or prefrontal dopamine transmission [99] , from mediodorsal striatum [101] , or from nucleus accumbens [48] impaired rats' performance on different dry-land or watermaze tasks requiring rapid place learning . Third , electrophysiological recording studies showed that neuronal activity in the hippocampus is coordinated with activity in medial prefrontal cortex [97 , 98] , mediodorsal striatum [100] , and nucleus accumbens [102] while rats are using rapidly encoded place information for efficient foraging behavior . Other subcortical sites , such as amygdala , lateral septum , and hypothalamus , which have been implicated in behavioral control and display strong connectivity to temporal and intermediate parts of the hippocampus [30 , 31] , may also interact with the hippocampus to translate place learning into behavior , but this possibility remains to be tested . The tendency , observed in the present study , for rats with hippocampal residuals at the temporal pole to show slightly better performance than those with a residual septal pole is consistent with the functional connectivity of the temporal 40% of hippocampus spared in our experiments; this area , apart from featuring strong links to prefrontal cortex and subcortical sites ( see Figure 1A ) , has some connectivity to the dorsolateral and adjacent intermediate domain of the entorhinal cortex [33] , and cells at the septal end of this area show relatively accurate place-related firing [38] . Although rats with only the septal pole of the hippocampus were markedly impaired on the behavioral tests requiring rapid place learning , their residual hippocampal circuitry exhibited intact entorhinal-hippocampal plasticity and could rapidly , within one exposure to a novel environment , form accurate and stable place-related firing in CA1 pyramidal cells . Our electrophysiological findings support the idea [39 , 57] that residual hippocampal circuitry at the septal pole , due to connectivity to the dorsolateral domain of the entorhinal cortex , can rapidly form and maintain accurate place representations , even in the absence of the rest of the hippocampus . However , such representations on their own are not enough to sustain task performance based on rapid place learning , as revealed by the poor performance of rats with only the septal hippocampal pole intact . The deficits in navigational performance could reflect that larger-scale place representations , normally provided by neurons in more temporal parts of the hippocampus , are required for route planning [18 , 37 , 40] . However , a selective deficit in route planning , i . e . , increased path lengths to reach the target area and relatively normal search preference for this area once it was reached , is not supported by our behavioral data . Rather , the marked performance deficits , despite an intact septal pole of the hippocampus maintaining accurate place-cell firing , highlight the importance of functional connectivity with prefrontal cortex and subcortical sites provided by the intermediate to temporal hippocampus . Notably , apart from “pure” place codes , normal hippocampal firing can show additional characteristics , including some that are indicative of the animal's goals or motivation [7 , 13 , 106–116] . Candidate neuroanatomical substrates to confer motivational information to hippocampal neurons include subcortical afferents from dopaminergic midbrain neurons and amygdala that mainly target the intermediate to temporal hippocampus [29 , 30 , 32 , 117–119] . Therefore , removal of the temporal to intermediate hippocampus may disconnect place cells in the septal pole from motivational modulation; such disruption might also contribute to performance impairments on the rapid place-learning task . Interestingly , the reduced between-trials stability of the peak firing rate we found in CA1 place cells from septal residuals ( experiment 4 ) may reflect a failure to accurately represent the procedural and motivational variables characterizing the recording trials ( e . g . , the rats had to perform a random-foraging task to receive food reward ) , given the ample evidence that the firing rate of place cells normally reflects such information [7 , 107–110 , 114–116] . Rats with a residual intermediate hippocampus performed better than all other partial lesion groups on all tests requiring rapid , one-trial , place learning . Remarkably , they generally performed as accurately and efficiently as the sham-operated control group . The exception to this was the retest on the rapid place-learning task in experiment 2 , when rats with a residual intermediate hippocampus were performing slightly worse than control rats , though still better than all other lesion groups . This deficit was likely related to interference from the preceding incremental training to one location . Possibly , throughout incremental training , hippocampal cells become increasingly recruited in the representation of the target location [111] , so that on a subsequent retest requiring rapid learning of a novel location , a reduced hippocampal circuit might have insufficient encoding capacity . Thus , it is important to note that even though a residual intermediate hippocampus can sustain performance based on rapid place learning , the septal and temporal tips are not redundant , but processing in these regions may be required to sustain normal behavior in more challenging situations . Nevertheless , the importance of the intermediate hippocampus for performance based on rapid place learning was further revealed by the finding that lesions removing this region , but sparing both the septal and temporal tip , completely abolished performance on the one-trial place-learning task , similar to complete hippocampal lesions . The intermediate hippocampus combines connectivity to the dorsolateral domain of the entorhinal cortex , where visuospatial information is represented at a fine scale by grid cells with fine-grained firing patterns—and consistently relatively accurate place-related firing can be recorded from the intermediate hippocampus [25 , 36 , 38]—with connectivity to prefrontal cortex and subcortical sites , including nucleus accumbens and mediodorsal striatum ( see Figure 1A ) . These connections are partially overlapping . In addition , extensive intrahippocampal projections [28 , 76–78] , which mediate excitatory transmission [120 , 121] , and synchronous neuronal activity [122] are found along the whole longitudinal extent of the intermediate hippocampus . Whereas the septal and temporal tip of the hippocampus , as spared in experiment 5 , should together also posses the two complementary sets of connectivity ( see Figure 1A ) , there are virtually no efficient connectional routes between the septal and temporal tip , except for those that involve steps in the intermediate region . Thus , the intermediate hippocampus anatomically and physiologically integrates two sets of functional links , namely to precise visuospatial processing , through the dorsolateral domain of the entorhinal cortex , and to behavioral control , through prefrontal cortex and subcortical sites , and our present data suggest that such integration is critical for the translation of rapid place learning into adaptive behavior . Whereas the present study highlights that the intermediate hippocampus is critical for the translation of rapid place encoding into behavior , previous functional imaging studies in humans and mice have suggested that the intermediate hippocampus also plays a key role in the encoding and retrieval of multimodal associative memory [123–125] . More specifically , information flow through the middle or intermediate hippocampus may mediate the binding or integration of different sensory modalities that , due to the topography of the relevant neocortical connections , are initially represented at distant septotemporal levels . This view cannot readily account for the importance of the intermediate hippocampus for performance based on rapid place learning , as our place-cell recordings indicate that the relevant place encoding and retrieval can be sustained by the septal pole without contributions by other parts of the hippocampus . However , it also emphasizes the integrative properties of the intermediate hippocampus . Overall , an intriguing picture is thus emerging of the intermediate hippocampus as a key substrate for functional integration along the longitudinal axis .
One hundred twenty-four adult male Lister-Hooded rats ( Charles River ) were used for the behavioral experiments ( experiments 1 , 2 , and 5 ) , 13 for the field-potential ( experiment 3 ) , and seven for the place-cell recordings ( experiment 4 ) . They were housed in groups of one to four in a temperature-controlled ( 20–23 °C ) and humidity-controlled ( 40%–55% ) room with an artificial light:dark cycle ( lights on: 7 am to 7 pm ) . Rats had free access to food and water . Only for the place-cell experiments were rats fed a restricted diet to maintain them at approximately 90% of their free-feeding weight , so that they were motivated to forage in the recording arena . Rats weighed 250–350 g and were 10–14 wk old at surgery . Before the start of experiments , all rats were habituated to handling by the experimenters . Experimental procedures were conducted during the light phase of the cycle as far as possible . Home Office regulations for animal experimentation were followed ( Project Licence No 60/2484 ) . Partial or complete fiber-sparing lesions of the hippocampus were made under halothane or isoflurane anesthesia using bilateral stereotaxic microinjections of the neurotoxin ibotenic acid ( Sigma; 10 mg/ml in 0 . 1 M phosphate-buffered saline ) through a 1-μl SGE syringe ( 26 ga , 0 . 47-mm-diameter needle; WPI ) , as described in detail elsewhere [68] , except that injection coordinates and volumes were adapted for the partial lesions to spare approximately 40% of hippocampal volume at different septotemporal levels . Partial or complete hippocampal lesions were achieved by injections of 0 . 05–0 . 10 μl at 7–13 sites in each hemisphere ( see Table S2 ) . For sham lesions , the empty injection needle was lowered into the neocortex at as many sites as required for the complete or partial hippocampal lesions , with the intention to induce comparable mechanical damage to the cortex . For the place-cell experiments , sham lesions or partial lesions intended to spare the septal 40% of hippocampus were combined with the implantation of one or two sets of four tetrodes ( four twisted , 17-μm polyimide-coated Pt-Ir[9:1] wires ) mounted in an inner cannula ( 28 ga ) connected to a light-weight microdrive ( Axona ) ; in addition , two rats without additional surgical treatment were implanted through small skull trepanations . The tetrodes were aimed above the CA1 layer in the septal pole of the hippocampus ( bregma and lambda horizontally aligned , 3 . 5 mm posterior and 2 . 2–2 . 5 mm lateral from bregma , and 2 . 00 mm below dura ) . An outer protecting cannula ( 18 ga ) on the microdrive was then lowered down to the dura , before Vaseline was applied around the cannula , and the trepanations above the lesions were filled with gauze for protection . The microdrive was secured to the skull using small screws , one of which served as electrical ground , and dental cement . Watermaze and general procedures . An open-field watermaze , 2 m in diameter and filled with water at 25 ± 1 °C made opaque by the addition of 200 μl of latex solution , was located in a well-lit room containing prominent visual cues at variable distance from the pool and visible from the water surface , so as to be used by the rats for orientation . To start a trial , rats were placed into the water , facing the pool walls , at one of four start positions ( north [N] , east [E] , south [S] , and west [W] ) . They could escape on a single platform , 12 cm in diameter and hidden from the animals' sight 1–2 cm below the water surface . The “Atlantis platform” was used , which can be withheld at >30 cm below the water surface for a predetermined time , by a computer-controlled electromagnet , before being raised to its normal position [126] . This enabled rewarded probe trials during which the rats' search preference is first monitored for 60 s , and the platform is then made available to reinforce spatially focused searching . Our paradigms involved testing with different platform locations on an inner ring ( 0 . 8-m diameter ) or outer ring ( 1 . 4 m ) concentric with the pool ( Figure S9 ) . Every trial ended with the rat sitting for 30 s on the platform before being returned to its cage . If an animal failed to reach the platform within 120 s , it was guided there by the experimenter . The rats' behavior was monitored by an overhead video camera connected to a video recorder and a computer with WaterMaze software ( Actimetrics ) in an adjacent control room . The software aided collection of the following measures: latency and path length to reach the platform location , swim speed , and the percentage time spent in the correct zone ( see Zone analysis of search preference below ) . Rapid place-learning task . A modification of the delayed-matching-to-place task [70] was used , in which trial 2 of a day was occasionally run as a probe trial . Rats received four trials a day . The platform was hidden in a novel location on trial 1 of each day and then remained in this place for trials 2–4 , on which rats could use rapidly encoded place memory to reach the escape platform efficiently . All four start positions were used daily in an arbitrary sequence , to discourage egocentric strategies . Analysis focused on trial 2 of each day , on which performance relied on place memory encoded within a single trial , whereas trial 3 and 4 were run to reinforce the win-stay rule of the task . Trials 1 and 2 were 10–30 s or 20 min apart ( i . e . , one-trial place memory on trial 2 was assessed at two different retention delays ) , whereas the delay was 10–30 s between the other trials . Search preference in the correct location on probe trials has long been recognized as the most reliable measure of place memory in the watermaze , whereas latency measurements are highly dependent on chance , and may be reduced efficiently through systematic search strategies and the use of single beacon cues and through a coarse estimate of position and sense of direction mediated partly by extrahippocampal structures [71] . Therefore , as an important modification of the original task , which purely relied on latency measurements [70] , trial 2 was occasionally run as rewarded probe trial to measure the rats' search preference for the zone containing the platform location . Incremental place-learning task . Rats were trained to a constant platform location with eight trials a day , ca . 10 min ( ±3 min ) apart , for 5 d ( i . e . , 40 trials altogether ) . Trials 2 and 6 of each day were run as rewarded probe trials . Start positions changed between trials , to discourage egocentric strategies . An additional rewarded probe trial was run on day 6 , to assess incrementally acquired memory unconfounded by within-day learning . Zone analysis of search preference . Search preference for the vicinity of the platform location on probe trials was assessed as follows: Eight 40-cm-diameter “virtual” zones , one of which ( the “correct zone” ) was concentric with the platform location , were defined on the inner and outer ring of the pool , so that the zones were nonoverlapping , evenly spaced , and symmetrically arranged ( Figure S9 ) . The time spent in each of these eight zones during the 60-s probe trial was determined automatically using the WaterMaze software . From these measures the “percentage of time spent in the correct zone” was calculated as: ( time in correct zone/time in all eight zones ) × 100% . By chance , i . e . , during random swimming , this value should be: 100%/8 = 12 . 5% , whereas higher values indicate a search preference for the correct zone . The fact that the comparison zones were located on the inner and outer rings was critical to unequivocally identify search preference due to one-trial place memory on the rapid place-learning task: throughout training , platform locations were always on the inner or outer ring , and hence , a search preference for a zone in this area as compared to a zone elsewhere on the pool surface might have merely reflected procedural and incremental learning . Experiment 1 . Three groups with partial hippocampal lesions sparing approximately 40% of hippocampal volume at the septal or temporal pole , or in the intermediate region , a sham-lesioned control group , and a group with complete hippocampal lesions were tested on the rapid place-learning task . Experiment 1 was run in five replications , each comprising 20 rats and always including sham-lesioned controls . Rats were pretrained on the rapid place-learning task for 8 d before the lesion surgery . After pretraining , rats were divided into groups , which were matched for all analyzed performance measures as far as possible , to receive the different lesions within 5 d . After at least 1 wk of recovery , 8 d of postsurgical testing on the rapid place-learning task commenced . Two sets of eight different platform locations ( one novel location per day ) were used during pretraining and postsurgical testing , with the eight locations evenly distributed across the pool ( Figure S9 ) . During both pretraining and postsurgical testing , the two delays between trials 1 and 2 ( 10–30 s or 20 min ) were used equally often , and on days 4 and 8 , trial 2 was run as probe trial . Two delays and two platform locations were used each day in all four possible combinations , so that the different delays were paired with the different locations in a counterbalanced manner across days with and without probe trials . The allocation of the different lesion groups to the different sequences of platform-delay combinations was counterbalanced . Experiment 2 . The lesion groups from experiment 1 were further tested on the incremental place-learning task , commencing 1–4 d after testing on the rapid place-learning task was completed . To avoid that results were confounded by the properties of one particular platform location ( rats tend to find some locations more easily than others ) , two platform locations were used , counterbalanced between groups , in distant parts of the pool . One location was on the outer ring in the N , the other one on the inner ring in the SW ( Figure S9 ) . Two days after testing on the incremental place-learning task , rats were retested for 1 d on the rapid place-learning task with a delay of 20 min between trials 1 and 2 . Rats trained to the location on the outer ring ( N ) were switched to the location on the inner ring ( SW ) and vice versa . Experiment 5 . Rats with lesions to the intermediate hippocampus , sparing approximately 20% at the septal and temporal tips , and a sham-lesioned control group were compared on the rapid place-learning task . The experiment was run in one replication including 12 rats in each group . Pre- and postsurgical testing was as described for experiment 1 . Field potentials from the dentate-gyrus granule cell layer in the septal hippocampus evoked by stimulation of the perforant path were recorded from anesthetized rats ( urethane , 1 . 5 g/kg intraperitoneally ) that had received sham lesions or partial lesions sparing the septal pole , followed by at least 1 wk of recovery . Electrode coordinates and procedures were similar to previous experiments [82] . The slope of the initial rising part ( 2 . 0–2 . 6 ms or 2 . 2–2 . 8 ms after stimulation ) and the population spike amplitude were used as measures of the evoked responses . After positioning the electrodes , low-frequency baseline stimulation ( biphasic 0 . 1-ms pulses , 0 . 05 Hz ) at 0 . 5 mA was applied until responses were stable . Recurrent feedback inhibition was tested by applying paired pulses , 20 ms apart , at 0 . 5 mA ( 15 pairs , 10 s apart ) [75] . To determine input–output curves , stimulation intensity was increased from 0 . 1 to 1 mA in 0 . 1-mA steps ( three stimulations per step , 0 . 1 Hz ) . To measure LTP , stimulus intensity was adjusted to obtain responses whose slope was approximately 50% of the slope maximum in the input–output curve , and low-frequency baseline stimulation ( biphasic 0 . 1-ms pulses , 0 . 05 Hz ) was applied for at least 20 min before tetanization . A tetanus consisted of three trains of 50 biphasic 0 . 2-ms pulses at 250 Hz with 60 s between trains ( overall 2 min ) . Following the tetanus , low-frequency baseline stimulation continued for an additional 60 min . At the end of the experiments , the locations of the electrode tips were marked by a 10-mA , 2-s biphasic pulse to the electrodes . The rats were then perfused and their brains further processed as described under Histology , below . Starting 4 to 7 d after tetrode implantation , rats were trained to forage in a 1-m × 1-m arena with a brown floor enclosed by four transparent Perspex walls ( 40-cm high ) and placed in a room with many distal cues . Water-soaked Rice Krispies were continuously scattered across the arena by the experimenter from behind a black curtain , so that the food-deprived rats moved continuously within the whole open field . Tetrodes were advanced daily to the CA1 pyramidal cell layer , and the rats became familiar with the random-foraging task and the environment . Neurophysiological signals from the tetrodes were recorded in parallel to the rat's path using Axona systems ( Axona ) . Once well-isolated hippocampal pyramidal cells were identified , experimental recordings began . Cells were recorded while the rats foraged for two ca . 15-min trials in the familiar environment ( Fam ) ( Figure 5B ) ; the trials were separated by typically 30 to 60 min , which the rat spent in a holding bucket and during which the arena was cleaned . Recordings in the familiar environment were repeated on subsequent days as the electrodes encountered new cells . In three rats ( one lesioned , two control ) , cells were also recorded during exposure to a new environment ( New ) to examine the rapid formation of place-related firing patterns . For this purpose , the familiar arena was replaced by a new 0 . 64-m × 0 . 94-m arena with a black floor and 3-cm lips , and two prominent distal visual cues were changed . The rats foraged during four ca . 15-min trials ( Fam-New-New-Fam ) , separated by ca . 10 min in the bucket ( Figure 5C ) . Spike sorting and subsequent detailed analysis of the firing characteristics and patterns of well-isolated pyramidal cells were performed offline with dedicated software ( for further details , see Text S1 , Supplementary Materials and Methods ) . In the watermaze experiments ( experiments 1 , 2 , and 5 ) , ANOVA ( with trials as within-subjects factor ) , followed by Fisher LSD post hoc tests , was used to analyze group differences . Paired two-tailed Student t-tests were used to compare latencies and path lengths on trials 1 and 2 , as well as the percentage time in the correct zone with the chance level , in individual groups . In the field-potential study ( experiment 3 ) , ANOVA was used to analyze input–output curves . To analyze LTP , slopes were averaged in 5-min blocks and expressed as a percentage of the mean slope during the 20-min baseline recording . Paired two-tailed Student t-tests were used to demonstrate LTP , and unpaired two-tailed t-tests were used for group comparisons . In the single-unit experiment ( experiment 4 ) , unpaired two-tailed Student t-tests were used to compare measures between the two groups . Accepted level of significance was p < 0 . 05 . After completion of the experiments , rats were anesthetized with an overdose of Euthatal ( Harlow ) and perfused transcardially with saline , followed by 4% formaldehyde solution . Brains were extracted from the skull , postfixed in 4% formaldehyde solution for at least 24 h , and then egg-embedded as described elsewhere [68] . Coronal 30-μm sections were cut on a freezing microtome , and every fifth section—or , for verification of electrode locations , every section—through the hippocampus was mounted on gelatine-coated slides and stained with cresyl-violet . For documentation , digital photographs were taken from exemplar brains . To measure the relative volume of spared hippocampal tissue , the intact hippocampal area ( dentate gyrus , CA1–3 fields ) on every fifth section was outlined at 10× magnification for each brain and measured using a binocular connected via a digital camera to a computer running Leica Q Win software . These areas were summed up for each brain , and the mean hippocampal area was calculated for each group . The relative hippocampal volume was calculated by dividing the hippocampal areas in individual brains by the mean hippocampal area determined for the relevant sham-lesioned control group . For the three-dimensional visualization and comparison of the distinct chunks of spared hippocampus in the different groups , Neurolucida and NeuroExplorer software ( MicroBrightField Europe ) on a computer connected to a microscope via a digital camera were used to prepare three-dimensional reconstructions of the hippocampus ( dentate gyrus , CA1–3 ) together with the brain silhouette from the coronal sections ( cut 150 μm apart ) of exemplar brains .
|
The ability to remember locations in space is dependent on an area of the brain called the hippocampus . A much-studied property of neurons in the hippocampus is that they rapidly come to represent or code for specific places—i . e . , the hippocampus “learns” places—as animals or humans move through an environment . Here , we identified in rats the hippocampal substrate enabling the translation of place learning into appropriate search and approach behavior ( similar to the task of returning to a novel place where you parked your car ) . We examined the impact of selective lesions to distinct parts of the hippocampus on behavior requiring rapid place learning and on in vivo electrophysiological models of hippocampal learning such as place-related neuronal activity . We showed that translation of rapid place learning into efficient search behavior requires the “intermediate” region of the hippocampus , a region that likely combines anatomical links to visuospatial information processed by the neocortex with links to behavioral control through prefrontal cortex and subcortical sites . In contrast , the so-called “septal” region of the hippocampus , which features the relevant anatomical links to visuospatial information processing , can sustain rapid place learning ( as reflected by formation of place-related neuronal firing ) , but not translate such learning into appropriate search and approach behavior .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"neuroscience",
"physiology"
] |
2009
|
From Rapid Place Learning to Behavioral Performance: A Key Role for the Intermediate Hippocampus
|
Cysteine residues have a rich chemistry and play a critical role in the catalytic activity of a plethora of enzymes . However , cysteines are susceptible to oxidation by Reactive Oxygen and Nitrogen Species , leading to a loss of their catalytic function . Therefore , cysteine oxidation is emerging as a relevant physiological regulatory mechanism . Formation of a cyclic sulfenyl amide residue at the active site of redox-regulated proteins has been proposed as a protection mechanism against irreversible oxidation as the sulfenyl amide intermediate has been identified in several proteins . However , how and why only some specific cysteine residues in particular proteins react to form this intermediate is still unknown . In the present work using in-silico based tools , we have identified a constrained conformation that accelerates sulfenyl amide formation . By means of combined MD and QM/MM calculation we show that this conformation positions the NH backbone towards the sulfenic acid and promotes the reaction to yield the sulfenyl amide intermediate , in one step with the concomitant release of a water molecule . Moreover , in a large subset of the proteins we found a conserved beta sheet-loop-helix motif , which is present across different protein folds , that is key for sulfenyl amide production as it promotes the previous formation of sulfenic acid . For catalytic activity , in several cases , proteins need the Cysteine to be in the cysteinate form , i . e . a low pKa Cys . We found that the conserved motif stabilizes the cysteinate by hydrogen bonding to several NH backbone moieties . As cysteinate is also more reactive toward ROS we propose that the sheet-loop-helix motif and the constraint conformation have been selected by evolution for proteins that need a reactive Cys protected from irreversible oxidation . Our results also highlight how fold conservation can be correlated to redox chemistry regulation of protein function .
Cysteine residues are involved in a plethora of roles in proteins , particularly in the context of cellular signaling , substrate and metal binding , protein–protein interactions and enzymatic activity . [1–4] However , most reactive cysteine residues are also quite sensitive to oxidative modification , leading to the formation of a diverse set of oxidized products when exposed to Reactive Nitrogen and/or Oxygen Species ( RNOS ) [1 , 2 , 5] . In this sense , oxidation is becoming an important regulatory mechanism in many proteins , and reactive cysteine residues are emerging as critical components in redox signaling [5] . A particularly oxidized form of cysteine , the sulfenic acid ( Cys-SOH ) , which has an important role as a sensor of oxidative and nitrosative stress in enzymes and transcriptional regulators , has a rich chemistry that can modulate the fate of protein activity . Sulfenic acid is a metastable oxidized form of Cysteine , which easily gives rise to more stable products like disulfides , sulfinic acid , or even sulfonic acids i . e . , overoxidation products [1 , 5] . Since reactive Cys oxidation usually leads to a loss of catalytic activity , there are several mechanisms that recover the reduced cysteine . These processes can be dependent on other proteins , small redox molecules ( like Glutathione ) , or they can even occur by an autorecovery mechanism promoted by the protein itself . One autorecovery mechanism depends on the formation of an intramolecular sulfenyl amide , a cyclic product that involves the reaction of the sulfur atom with the backbone NH moiety of the succeeding residue protecting it from overoxidation . Other autorecovery mechanisms involve the reduction of the oxidized cysteine with a nearby cysteine residue to produce a disulfide bond [6] . All these possible Cys oxidation/reduction reactions are shown in Fig . 1 . The sulfenyl amide intermediate was first observed in the crystal structure of human protein tyrosine phosphatase 1B ( PTP1B ) [7] . Protein tyrosine phosphatases regulate signal transduction pathways involving tyrosine phosphorylation and have been implicated in the development of hypertension [8 , 9] , diabetes [10–12] , rheumatoid arthritis [13–16] and cancer [17–21] . Increasing evidence suggests that the cellular redox states of the catalytic cysteine are involved in determining tyrosine phosphatase activity through the reversible oxidization of the reactive cysteine to sulphenic acid ( Cys-SOH ) [18 , 22–24] . Hydrogen peroxide ( H2O2 ) can regulate cellular processes by the transient inhibition of protein tyrosine phosphatases through the reversible oxidization of their catalytic cysteine , which suppresses protein dephosphorylation [25–27] . In this sense , the discovery of sulfenyl amide formation in PTP1B emerged as a possible mechanism to recover a functional reduced cysteine . Since the discovery of the formation of the sulfenyl amide intermediate in PTP1B [7] , several other proteins have been identified to harbor this intermediate , thus showing that sulfenyl amide formation is emerging as a common post-translational modification , related to protein redox reactivity . For example , in Bacillus subtilis OhrR , it is involved in the control of peroxiredoxins expression in response to ROS . Cyclic sulfenyl amide prevents the overoxidation of this repressor , and acts as a slow switch to prevent DNA binding , allowing the transcription of the peroxiredoxin genes [28 , 29] . Another protein where cyclic sulfenyl amide was detected is PTPalpha , composed by two domains , one proximal ( D1 ) , which has phosphatase activity and one distal ( D2 ) , which is not directly involved in phosphatase activity . The cyclic sulfenyl amide has been described in the D2 distal domain; in this case it functions as an allosteric regulator of the D1 domain , controlling its catalytic activity [24] . From another perspective , cysteine residues are highly reactive towards RNOS and several works have shown that protein environment regulates this reactivity by controlling not only the interaction with the oxidative species , but also by modifying for example the pKa of the thiol group [30 , 31] . Interestingly , once the first oxidation has occurred yielding the corresponding sulfenic acid , there is little information about the molecular determinants that regulate its fate . The oxidized cysteine can follow one of two possible paths; to form a cyclic sulfenyl amide , which can then be recovered; or to be irreversibly oxidized to sulfinic or sulfonic acid . ( Fig . 1 ) The reaction of sulfenic acid to form sulfenyl amide has been previously studied in model compounds suggesting that electronic effects are relevant but they report a high energy barrier . [32 , 33] In the present work we have studied the key structural and chemical features that allow sulfenic acid to form the sulfenyl amide intermediate in proteins using in silico based tools . Our results show that: I ) a specific and conserved beta-sheet-loop-helix motif is present across different protein folds , and positions the NH backbone , which reacts with the cysteine sulfur atom to yield the sulfenyl amide , in a constrained conformation required for the chemical reaction to occur . II ) The intramolecular reaction occurs , in a concerted fashion , with S-OH bond breakage as the rate limiting process . III ) Protein families having the constrained cysteines motif are reported to be involved in redox related process , strongly suggesting a functional relationship . IV ) A database search for the motif shows its presence in other proteins , like the protein tyrosine phosphatase B from Mycobacterium tuberculosis , suggesting their possible role in redox related signaling .
In order to work with all available structures deposited in the Protein Data Bank , a relational database was built using MySQL [34] as the backend . This database stores information such as the UniProt ID , PFAM family ( computed by HMMer [35] ) , primary , secondary and tertiary structural data like protein sequence , secondary structure ( such as Alpha Helix or Beta sheet , computed by DSSP [36] ) , aminoacid-aminoacid contacts and phi/psi dihedral angles for every aminoacid . All these data can be used as search parameters in the database . For the current analysis , all unique proteins ( as defined by UniProt ID [37] with a structure deposited in the PDB [38] ) were considered . The use of UniProt Id significantly reduces the redundancy of the PDB but does not eliminate the bias due to differential representation of protein families or multiple structures of highly similar proteins . For this reason we applied also a sequence similarity filter , i . e . , we considered only one structure for all sequences with over 95% identity . The total number of structures used in our study turns out to be 18 , 547 . We removed all entries corresponding to short peptides , fully unstructured regions or those near to unresolved zones in the X-ray structure . NMR structures were selected only when the conformation was represented in at least 50% of the reported conformational modes . For the evaluation cysteine conformation , all crystal structures depicted above were filtered considering only proteins that have a cysteine residue whose psi dihedral angle is between -150 and -90 degrees ( i . e . in the forbidden psi conformation ) . We also filtered crystals in which the constrained cysteine is involved in disulfide bonds and crystal structures with resolution of 2 . 5 Å or higher . This protein selection pipeline is described in S1 Fig . We found , and visually analyzed case by case , 145 proteins showing the presence of a Cys in the forbidden-psi conformation and the helix-beta-loop-helix motif . The schematic representation of each family's secondary structure , shown in Fig . 2 , has been taken from the PDBsum website and done with HERA—A [39] . Analysis of secondary structure around Cys was performed using DSSP[36] . Tertiary structure analysis was performed by directly computing several structural parameters from the corresponding PBDs . In order to analyze the contribution of primary , secondary or tertiary structure to the stabilization of the Cys forbidden-psi conformation we performed two different strategies due to the fact that involved proteins belong to different families and are dissimilar in their sequences . A first approach was done by multiple sequence alignments ( MSA ) that were built harboring the Cysteine residue with 20 flanking residues at either side of the Cys . As a second approach , a structural alignment was computed with the whole secondary structure motif from each PDB structure , using the SALIGN algorithm [40] . The protein sequence for both alignments was done by choosing only one centroid sequence along a 95% clustering computed using CD-HIT [41] . Hidden Markov Models ( HMM ) were built using previously mentioned MSAs with HMMMER [42] . Each HMM was tested for their capacity to detect proteins harboring the forbidden-psi Cys . For visual analysis , HMM and frequency logos were built using Skylign [43] . Molecular Dynamics Simulations ( MD ) . The starting structure for the MD simulations was retrieved from the Protein Data Bank [38] , corresponding to PTP1B with Cys215 in the sulfenic acid state , ( PDBid 1OET ) with a crystal resolution of 2 . 3 Å . Using this structure three different systems were built varying in the protonation state of Histidine 214 , which was either protonated in the delta Nitrogen ( HID state ) , in the epsilon Nitrogen ( HIE state ) or in both nitrogens , resulting in a positively charged Histidine residue ( HIP state ) . Standard protonation states were assigned to all other titrable residues , D and Q were negatively charged , K and R positively charged and Histidine protonation was assigned favoring formation of hydrogen bonds in the crystal structure , but in the case of the already mentioned Histidine 214 . For the peptide simulations we built a small molecule containing the sulfenic acid and both the anterior and posterior peptide bonds , capped with acetamide ( ACE ) and N-methyl ( NME ) groups respectively , consisting of 24 atoms ( See Fig . 3C ) . For all residues , except the sulfenic acid , the AMBER99SB force field was used [44 , 45] . Sulfenic acid force field parameters were built using AMBER recommended procedure . Briefly , an electronic structure calculation using the HF/6–31G* method was performed , and partial atomic charges were subsequently derived using RESP procedure[46] . All bonded and VdW parameters were taken from the General AMBER Force Field[47] . Parameters for the resulting cysteine displaying a sulfenic acid side chain are shown in S2 Fig . Each protein was then immersed in a truncated octahedral box of TIP3P water consisting in 8 , 776 water molecules , which corresponds to a 10 Å distance between the protein surface and the box boundary [48] . Each system was optimized using a conjugate gradient algorithm for 2000 steps . This optimization was followed by 100 ps long constant volume MD , where the temperature of the system was slowly raised to 300 K . The heating was followed by a 100 ps long constant temperature and constant pressure MD simulation to equilibrate the system's density . During these processes the protein Cα atoms were restrained by a 1 kcal/mol harmonic restraint potential . Pressure and temperature were kept constant with the Berendsen barostat and thermostat respectively adjusting pressure every 1 ps . and temperature every 2 ps , using the Amber suggested default parameters . [49] All simulations were performed with periodic boundary conditions[50] using the SHAKE algorithm[51] to keep hydrogens at equilibrium bond lengths , and using a 2fs time step . Production simulations consisted in 10 ns long NTP simulations for the protein , and 100ns for the model peptide . Ewald sums were used to treat long range electrostatics , using AMBER default parameters , and with a 10 Å cutoff for direct interactions Constant pH molecular dynamics ( CpHMD ) . For constant pH molecular dynamics , unless explicitly stated , simulation parameters were the same as detailed above . A detailed description of the parameters is presented in the original paper of CpHMD simulations [52] . Simulations were done with the Generalized Born implicit solvent representation [53] . A cutoff of 1000 Å was used for direct interactions . Temperature was kept constant ( 300 K ) using Langevin dynamics with a collision frequency of 2 . 0 fs-1 . [54] Each CpHMD simulation consisted of 10ns . In order to compute the pKa , fitting to the Henderson-Hasselbalch equation was performed using non linear least square algorithm as implemented in R 3 . 0 package . [55] .
Graphics were produced with R 3 . 0 statistical package[55] . VMD has been used to produce images of the protein or the peptide[71] .
As mentioned in the introduction , PTP1B has been crystallized with Cys215 in both the oxidized , sulfenic acid state , as well as the cyclic sulfenyl amide state . A closer look at this residue sequence and structural environment shows one interesting observation . Cys215 displays , both in the reduced and sulfenic acid states , a position in the Ramachandran plot which usually constitutes a forbidden zone . This conformation results ( as shown in Fig . 3A ) in a configuration that orients or directs the side chain of the Cys residue in the same direction as the NH hydrogen of the Cysteine215-Serine216 peptide bond ( i . e the NH of the following residue ) , which is the nitrogen required to form the cyclic sulfenyl amide . We will call this cysteine conformation the “forbidden-psi” conformation or constrained conformation . We also analyzed whether there are any other proteins displaying cyclic sulfenyl amide in the PDB . Apart from PTP1B , we found only one case , Phospho-2-dehydro-3-deoxyheptonate aldolase AroG from Mycobacterium tuberculosis ( S1 Table ) , which seems to harbor a cyclic sulfenyl amide ( S-N distance is 1 . 85A ) . In AroG the cysteine residue is located in a long unstructured loop with no clear catalytic function described . Moreover , the presence of the sulfenyl amide is not mentioned by the authors[72] . In order to analyze how often a Cysteine residue is found adopting the corresponding forbidden-psi conformation , we surveyed all PDB structures[38] and measured their psi and phi angles for all Cysteines . The resulting 2-dimensional histogram is shown in Fig . 3B . As expected , most Cys residues are found in the allowed zones ( corresponding to alpha and beta secondary structures ) . However , there is a significant number of Cys residues displaying forbidden-psi values , i . e . falling in the zone delimited by the red lines in Fig . 3B . To characterize them further , we selected all unique protein structures ( as defined in methods ) with a Cys adopting the forbidden-psi value ( corresponding to a range of psi values between -150 and -90 degrees , resulting in 270 structures ( See S2 Table for a full list of the corresponding PDB entries ) . As an example , Fig . 3C and 3D show respectively an oxidized sulfenic acid Cys residue adopting a common beta structure conformation and the forbidden-psi conformation . Structural characteristics of the forbidden-psi Cys . To analyze the structural surrounding of the relevant Cys , we used two different approaches . First , we characterized the immediate environment of the forbidden-psi Cys , by selecting all residues having at least one atom less than 8 Å away from the cysteine center of mass . However , we could not identify any over represented aminoacid ( or aminoacid type ) or any conserved set of interactions , not even Histidine , a residue that was proposed to form a hydrogen bond with the carbonyl group of the cysteine peptide bond in PTP1B and relevant for Cys reactivity . Secondly , we thought about the possibility of the local protein fold being responsible for forcing the Cys to assume the forbidden-psi conformation . Strikingly , we found 53% ( 145/270 ) of the proteins displaying a forbidden-psi Cys adopt the same local fold around it , characterized by a beta strand-loop-helix secondary structural element with the relevant Cys located at the end of the beta-strand , which is also part of a parallel beta sheet motif , with at least three strands . Moreover , the Cys containing strand appeared to be always the one in the center of the three beta-strands . The corresponding fold is shown in Fig . 2 . The PFAM families were checked in order to see if the proteins having a Cys displaying the forbidden psi conformation but lacking the structural motif correspond to common protein functions . Only two families have a significant number of structures , more than three unique proteins , with a constrained cysteine: Retroviral aspartyl protease ( PF00077 ) and Beta-lactamase2 ( PF13354 ) ( S3 Table ) . In the Retroviral aspartyl protease family , the cysteine in the forbidden psi conformation is in a turn between two beta sheets , thus a similar motif . Crystal structures of this protein family are generally homo-dimers , with a subunit presenting the Cys in the forbidden-psi conformation , while in the other one adopts a left handed helix conformation . Overoxidation of the cysteine residue has not been reported in this family . In the case of the Beta-lactamase2 family , many structures present a disulphide bond between the constrained Cys and a nearby Cys placed in a beta sheet . As in the case of Dual specificity phosphatases like PTEN , it is possible that disulphide bond formation is faster than sulfenyl amide formation . Overoxidation of these cysteine residues has also not been reported . All the other proteins identified with the constrained cysteine belong to families with only one structure . Taking this into account , from now on we concentrate our analysis in the 145 proteins that have the forbidden Cys and also have the same local fold . Sequence characteristics of the forbidden-psi Cys . Initially , we looked for any conservation in the sequence surrounding the forbidden-psi Cys by analyzing two different length segments , one corresponding to 20 residues at each side of the Cys and another including the whole secondary structure motif harboring the Cys residue . We performed multiple sequence alignment ( MSA ) and built the corresponding HMM either fixing the alignment without gaps around the Cys for the short segments or performing structural alignment for the whole motif . We used the built HMMs to detect those proteins sequences harboring the forbidden-psi Cys in the whole SWISS-PROT sequence database [73] . The results are shown in S4 Fig . The fixed model is able to find 93 out of 145 proteins with the motif ( 64% ) whereas the Structural model only recognizes 40% of the 145 proteins . Interestingly , the search also retrieves proteins ( whose structure is unknown ) with Phosphatase activity and GATase activity , which presumably could display a forbidden-psi Cys and/or the structural motif . The search also retrieves some false positives . Visual analysis of the HMM logo ( S5 Fig . ) shows some partially conserved residues like a His residue before the Cys , a rather conserved Glycine three residues ahead and an Arginine also rather conserved six residues ahead . Family assignment and analysis of the proteins containing the forbidden-psi Cys Fold . Having identified a common structural fold around the forbidden-psi Cys , we looked at how this element is inserted in larger protein folds or domains . For this sake , we assigned all found structures to PFAM families[74] . Interestingly , most structures ( 104 out of 145 structures , 72% ) with the forbidden-psi Cys are found in only seven protein families . Given that PFAM families usually define unique protein structural and functional domains , we analyzed how many of the reported structures from each family have a forbidden-psi Cys . As expected , most of the solved structures display the forbidden-psi . Remarkably , as shown in Fig . 2 , global folds corresponding to the families harboring the reactive Cys are quite different , despite having the conserved forbidden-psi local fold . We identified two big families of proteins , phosphatases ( with three PFAM families ) and glutamine amido transferase ( with two PFAM families ) . A structural alignment of the structural motif is shown in S6 Fig . These results are summarized in Table 1 . Assignment of the forbidden-psi Cys containing proteins to families , prompted us to explore whether these proteins were reported to play a role in oxidative processes , and thus gain some insight on the likelihood that the cysteine , its sulfenic acid and/or cyclic sulfenyl amide , could be physiologically relevant . For this sake , we performed a systematic literature search for any information related to Cys oxidation in each of the relevant families reported in Table 1 . Surprisingly , for five out of the seven families , we found reports relating the forbidden-psi Cys with either catalysis or a regulatory role , and a specific mention to a directly related oxidative process . ( Table 1 and references therein ) . We now will comment on these families ( Specific proteins with relevant data are presented in S4 Table ) : Protein Tyrosine Phosphatase ( PF00102 , Y_phosphatase ) . As commented above , PTPs are involved in a plethora of biological processes and are sensitive to oxidative stress . In this PFAM family 33 proteins have been found with the forbidden-psi Cys . The cysteine residue in the forbidden region is involved in the catalytic activity of these proteins and has been shown to be oxidized to sulfenic acid and to form cyclic sulfenyl amides ( The already mentioned human PTP1B belongs to this group ) [7] . Glutamine amidotransferase ( PF00117-GATase ) . Proteins from this group are involved in the transfer of the ammonia group of glutamine to an organic molecule . Detected Cysteine residues belong to the catalytic triad of these enzymes[80] . In analysis of the 18 unique proteins crystallized from this family all 18 have the constrained cysteine . Nevertheless , oxidation has not been observed in any of the crystallized proteins . In this sense , we foresee that redox agents could regulate proteins from this family , as they have the “constrained conformation” , the conserved motif , and a relatively exposed Cys residue . Rhodanese-like domain ( PF00581 ) . Members of this family include Cdc25 phosphatase catalytic domain , non-catalytic domains of eukaryotic dual-specificity MAPK-phosphatases , non-catalytic domains of yeast PTP-type MAPK-phosphatases and many bacterial cold-shock and phage-shock proteins . The cysteine residue is involved in catalysis and has been described in its oxidized state[81] ( as sulfenic acid ) . In this case 92% crystallized proteins have the constrained cysteine . Dual specificity phosphatase catalytic domain ( PF00782 ) . These proteins are able to dephosphorylate proteins with both pTyr and pSer/pThr residues and a cysteine residue is involved in the reaction . Oxidation of the reactive cysteine has been observed in some of its members[76] . In this case , 95% proteins with crystal structure have the constrained cysteine . Carbon-nitrogen hydrolase ( PF00795 ) . These enzymes are involved in the breakage of a carbon-nitrogen bond in different compounds . Again , this group of proteins have a catalytic cysteine[82] involved in the reaction . Although oxidation of these cysteine residues has not been reported yet , all of the proteins have cysteines in the unfavorable region . SNO glutamine amidotransferase ( PF01174 ) . Members of this family are involved in the biosynthetic pathway of vitamin B6 ( Pyridoxal phosphate ) and are active in its hetero oligomer state . This oligomer is formed in an equimolar relationship of one amidotransferase chain ( called Pdx2 ) and one synthase domain ( called Pdx1 ) . [83 , 84] Oxidation of the catalytic cysteine has been reported for pdxT from Staphylococcus aureus . [77] Only one member of this group does not have the cysteine contraint conformation . DJ-1/PfpI ( PF01965 ) . Proteins from this family include transcriptional regulators , proteases , chaperones and proteins with diverse roles such as DJ-1 which is involved in the development of Parkinson's disease . Because of its pathological relevance and protective role in oxidative stress DJ-1 has been intensively studied and oxidation of the active site cysteine has been described several times [78 , 85] . All the proteins from this group have the constrained cysteine . In summary , global analysis of all available unique protein structures shows that there is a significant number of them harboring a Cys residue displaying a conformation with the psi angle in a forbidden region ( -90° to-150° degrees ) , that orients the Cys side chain in the same direction as the next peptide bond NH moiety . Unexpectedly , structural domain analysis shows that the forbidden-psi Cys is in a large number of cases located in a motif consisting of a strand- ( Cys ) -loop-helix motif , inserted in several different global protein folds . They correspond to , at least , seven different protein families ( according to PFAM ) in which the Cys residue is important for catalysis and for five of these families . There have been reports on cysteine oxidation to sulfenic acid , implying that redox regulation may be associated with our findings . The pKa of the Cys with the forbidden-psi . Cys reactivity is tightly related to its pKa . In particular , Cys oxidation is promoted for those Cys with lower pKas which display significant population of the charged state . Therefore , we decided to analyze whether forbid1den-psi conformation and secondary motif could affect it . We used constant pH MD simulations to determine Cys pKa in both a constrained model peptide in the forbidden-psi conformation and a small peptide harboring the whole secondary structure motif taken from the crystal structure of PTP1B . The results show , as expected , that in the free peptide the reference pKa for Cys is obtained ( 8 . 50 ) . Imposing a psi angle restriction results in a slightly higher pKa ( 9 . 04 ) , a difference within the order of the error of the method which indicates that the constrained psi-conformation is not inducing a change in the pKa for the Cys . Interestingly , in the case of the peptide mimicking the structural motif , the computed pKa value is 4 . 82 . Thus , it is clear that the secondary structural motif lowers the pKa of the active cysteine . The extreme low pKa could be an artifact which allows to take only part of a protein and to highlight the role that the local structure plays lowering the pKa . We also decided to analyze Cys pKa in PTP1B , which is our test case . Excitingly , CpHMD simulations show that in PTP1B the Cys protonation state is coupled to a small but significant conformational change that results in Cys displaying a conformational dependent pKa yielding extreme values of 0 and 11 . 5 . The unusually low value seems to be the result of several strong hydrogen bond interactions that the deprotonated Cys performs with the protein environment ( Shown in S7 Fig . ) . Although in these cases obtaining the pKa requires knowledge of the conformational equilibrium constant , previous experimental estimations yielded a value of 5 . 6 [86] , which again shows that reactive Cys pKa is lowered . We now turn our attention to the chemical reaction of forbidden-psi Cys in the sulfenic acid state to yield cyclic sulfenyl amide , using human PTP1B as a test case . Our hypothesis is that the forbidden-psi conformation is directly responsible for the formation of cyclic sulfenyl amide . Energetic analysis of the forbidden-psi Cys conformation in a model peptide . The results presented above highlight the relationship between the forbidden-psi conformation and the conserved beta strand-loop-helix motif with the functional relevance of Cys residues and its possible implication in redox regulation . We initially analyzed the free energy difference between the forbidden-psi conformation and allowed helix conformation . The data presented in Fig . 3B allows an estimation of how much energy proteins must pay to constraint the Cys in the reactive ( forbidden-psi ) conformation using the Ramachandran plot derived free energy , estimated it around 5 . 5 kcal/mol . We then conducted an independent estimation of the corresponding cost in the sulfenic acid form . For this sake , we built a small peptide containing a sulfenic acid oxidized cysteine capped with Acetyl and N-Methyl groups , in the N and C terminal respectively ( as shown in Fig . 3C and 3D ) . We then performed 100ns long MD simulations for the peptide containing Cys-OH in water . The MD results ( shown in S8 Fig . ) , show that rotation along the psi dihedral angle has two minima , one at -30 degrees , spanning from 60° to -20° corresponding to helix like structures , and a second one with the minimum at 150° , spanning from 120° to 180° corresponding to structures in a sheet-like conformation . Interestingly , the peptide presents almost no conformations in the -60° to -160° range , during the whole simulation time scale . Free energy estimations show that the “forbidden psi conformations” are over 5 kcal/mol higher than the two minima , in agreement with the previous Ramachandran plot analysis and the results of Hornak et al for Ala tetrapeptides , where this region of the Ramachandran plot has a free energy higher than 5 kcal/mol [44] . Clearly , our results show that the protein must pay a considerable ( free ) energy cost to have a cysteine in the reactive or forbidden-psi structure , both in the Cys and sulfenic acid form . Since potential SOH to backbone amide interaction could stabilize the constrained conformation , we analyzed the likelihood of internal hydrogen bond interactions between the amide hydrogen and either the sulfenic acid S or O atoms . Distances and angle measurement during the simulation show that there is not a strong interaction that could be accounted as a hydrogen bond during the simulation timescale ( i . e . HNH-S and HNH-OSOH ) distances are larger than 3 . 5 Å most of the time ) ( S3 Fig . ) . Protein environment effects on Cys conformation in PTPB1 . Environmental structural analysis revealed that there are not clear interactions around the Cys residues that could be favoring not only the constrained conformation but also the sulfenyl amide formation . However , as shown by S3 Fig . His214 ( depending on its tautomeric state , see below ) may establish a hydrogen bond with Cys215 carbonyl , an interaction which has been suggested to increase the partial charge on the backbone nitrogen enhancing its reactivity and supporting a nucleophilic substitution mechanism for PTP1B[87 , 88] . Taking this into account we decided to analyze the role of the Histidine tautomeric state . In order to analyze the cyclic sulfenyl amide reaction mechanism in PTP1B ( see below ) and the role played by His214 ( in all possible tautomeric states ) we performed 10ns long MD simulations starting from the Cys215-SOH modified PTP1B setting histidine tautomeric states either as HIE , HID or HIP ( see Methods for details ) . The results show that the protein is stable in all three systems but significant differences are observed concerning the local structure of the Cys215 loop . S3E Fig . shows histograms for the Cys215 psi angle for all three states . As shown by the figure it is clear that Histidine protonation state affects Cys psi dihedral angle . When His214 is in the HIE state , no hydrogen bond interaction can be established and as consequence the psi angle shows values further from 180° ( mean value is 126° ) . When His214 is simulated as HID , hydrogen bond between His214 and Cys215 carbonyl forms and breaks several times during the simulation ( see S3B Fig . ) with a population of ca 50% . Consequently , the Cys215 psi angle has an average value of 175° , whereas when His214 is protonated ( HIP state ) the His214-Cys215 hydrogen bond is present 90% of the time ( See S3B Fig . ) and the average psi value is -165 degrees . In order to have an estimation of each His tautomer population , we performed constant pH MD simulations using His214 as the titrable residue . The results , show that at pH = 7 HID is the most populated state , and pKa is estimated to be around 4 . These results show clear evidence linking the His protonation state with the Cys-SOH conformation . Being HID the most populated state at physiological pH and HIP the one which enhances the forbidden-psi conformation; we decided to perform QM/MM MSMD computations with PTP1B His 214 in both HID and HIP states QM/MM study of sulfenyl amide formation reaction mechanism . In order to understand in detail the reaction mechanism of cyclic sulfenyl amide formation , we determined the corresponding free energy profile using a QM/MM strategy as explained in methods . The easiest reaction mechanism that can be envisaged requires the Cys-OH group to take a hydrogen atom or a proton from the backbone amide group of the previous residue ( Ser216 in this case ) to form the leaving water , leading to subsequent N-S bond formation . This mechanism has been tested in model systems by other groups [32] giving activation barriers ca . 50 kcal/mol , thus too high to account for a biological relevant process . Indeed we obtained similar values for the reaction using a model peptide in vacuum ( See S9B Fig . ) . Therefore , we thought on possible alternative mechanisms . In proteins , the reaction occurs in water , and since the key event in the reaction seems to be the breakage of the S-O bond , we decided to test whether the presence of explicit waters in the QM system could yield smaller barriers . To test this idea , we included in the QM system 10 water molecules and explicitly promoted proton transfer from the solvent to the S-OH group . The results ( presented in Table 2 and Fig . 4 ) show that the presence of explicit waters is key for determining the reaction mechanism and barrier . The free energy barrier is 13 . 9 kcal/mol ( Fig . 4A ) , which yields an intermediate with a broken S-O bond and a well formed S-N bond , but the N is still attached to the amide proton , thus having sp3 like character . In a second step , the amide proton is released to water , almost barrierless , yielding the cyclic sulfenyl amide product . The reaction is moderately exergonic by ca -14 kcal/mol . Distance analyses along the reaction ( Fig . 4B ) , show that the first step occurs in a concerted fashion , as soon as the S-OH bond is broken ( red line ) the S-N bond forms ( black line ) and this process occurs simultaneously with proton transfer from the solvent to the S-OH group ( green and yellow lines ) . The TS depicted in Fig . 4 shows a completely broken S-O bond , a well formed water molecule and the S and N atoms quite close at a distance of 1 . 93 angstroms . After the TS the key event is proton transfer from the NH to the solvent ( blue line ) . During the reaction the leaving Oxygen increases its negative charge , while the NH proton slightly increases it . Also , as expected , along the reaction the psi dihedral angle does not change significantly , until the end of the reaction reaching a value of -150° . The fact that the first and most important TS requires water release after proton transfer from the solvent , suggests that the reaction first step rate may be enhanced in acidic media . To analyze possible pH effect , we also computed the reaction free energy adding one hydronium ion hydrogen bonded to the S-OH group in the QM system . The resulting FEP and mechanistic analysis shows that reaction proceeds similarly as described above , but the barrier is slightly smaller 10 . 6 kcal/mol . This slight decrease in the barrier is due to the fact that transferring the proton from the hydronium ion is easier than from water . Lastly , we also computed the free energy setting His214 in the HIP state and using a hydronium ion ( S10 Fig . , blue curve ) , again mechanistic analysis shows similar results and the barrier is similar as in the previous case , thus His protonation state does not seem to affect significantly the reaction barrier . In summary , despite the second step is expected to decrease its rate when lowering the pH since the solvent must act as a base . Given the above mentioned results , and since the first barrier is significantly larger than the second , pH effects are expected to affect each barrier differently and possibly enhance the reaction rate . In order to analyze whether the protein environment and the conformational restraining effect , we performed the reaction in a model peptide in a box of waters . Interestingly , the reaction occurs with a similar mechanism and with almost the same barrier as in the protein , but only if the peptide conformation is restrained to the forbidden-psi angle ( See S9C Fig . ) . Trying to make the reaction to happen with Cys in a non forbidden conformation results in non reactive trajectories . As we stated in the methods section , to determine the accuracy of the level of theory used to compute the free energy profiles , we determined the reaction by using the Hybrid program [65 , 66] . Similar results were obtained with an activation barrier of 9kcal/mol ( S9C Fig . ) , showing that DFTB yields good results and can be used with free energy scheme . Analysis of the charge ( Table 2 ) of the involved atoms during the reaction shows that most of the atoms do not have a relevant change in their atomic charge . We observe only an increase in the Os atom that is due to its transfer from the sulfenic molecule to form a water molecule . There is also a slight decrease in the backbone nitrogen , as it binds the sulfur atom but keeps the hydrogen that is partially restored once the hydrogen is transferred to the solvent . In summary , our results show that the reaction mechanism involves proton transfer from and back to the solvent , with the heterolyic breaking of the S-O bond and formation of the leaving water as a key process . The reaction has a moderate barrier and thus is expected to occur readily . Clearly , neglecting the presence of explicit waters , as in previous works [32 , 33] yielded barriers which are too high to be compatible with a physiological role . Product structure . An important point should also be made concerning cyclic sulfenyl amide product structure and the Cys psi- dihedral angle . The analysis is similar to that of the phi-values of any Proline residue , due to the intra residue N-C bond . Briefly , given the non aromatic characteristic of the Cα and Cβ atoms of the five membered ring , the cyclic structure is non planar as shown in Fig . 5 . As already discussed , the key parameter for the reaction is the psi angle , which involves rotation along the Cys Cα-C bond , and which in turn defines the relative orientation of the residue side chain , including Cβ . As a consequence , fixing the Cβ position in the heterocycle as in the product imposes a strong constraint in the Cys psi-angle . Our results show that using the Cys-Ser peptide bond plane as a reference , which also contains the Cys Cα ( Dashed line in Fig . 5 ) , the Cα-Cβ bond can be positioned establishing a ca . 20° to 30° angle to either side of this defined plane , as shown in Fig . 5B and C . As a result , when the angle is negative ( counterclockwise ) a psi angle of ca -155° is imposed to Cys , while for positive angles ( Fig . 5C ) the imposed Cys psi angle is ca . -105° . These results confirm that if the protein Cys cannot adopt any of the mentioned “forbidden psi” values , cyclic sulfenyl amide formation is impossible . Taking all results together , it is clear that the reaction barrier is low , that the mechanism is clearly dissociative , and that there is no role for the protein in catalysis but to position the Cys psi angle in the constrained but reactive conformation compatible with the cyclic product structure .
In this work we have shown that cysteine reactivity can be controlled by the protein topology thus acquiring a specific conformation that regulates the barrier to form cyclic sulfenyl amides . ( Fig . 6 ) We started our analyses by identifying the presence of a Cys residue displaying a forbidden-psi angle in the -90° to -150° range in PTP1B known to form cyclic sulfenyl amide , and therefore performed a search across all protein structures found in the PDB . We were able to identify a set of protein families that have a significant number of members with the constrained cysteine that are involved in redox processes . Moreover , the identified proteins share a common topology that seems to be relevant for lowering the reactive Cysteine pKa and therefore enhancing their catalytic activity . However , this also enhances Cysteine reactivity towards ROS , and inactivation of the proteins . According to our study , it seems that this motif has been selected by evolution to accelerate catalytic activity and also to protect the cysteine from further oxidation , once is oxidized to sulfenic acid , by catalyzing the formation of a cyclic sulfenyl amide that can then be recovered to cysteine . We identified seven PFAM protein families with several members with the conserved structural motif as we pointed out before . The most important in terms of available experimental information is the protein tyrosine phosphatase family where the first cyclic sulfenyl amide was identified in the crystal structure of PTP1B[87] . This cyclic sulfenyl amide product has also been described in PTPAlpha after H2O2 treatment of the protein [24] . However , for some members of this family like the SH2 phosphatases , which have a constrained conformation and a conserved topology , some reports have detected disulfide bonds instead of sulfenyl amide [23] . Similar results have been published for Cdc25 , Rhodanase-Like domain , [26] , and for PTEN of the Dual specificity Phoshpatase domain[89] . Interestingly , all these proteins , and not other members of the family , have another Cys in the vicinity of the constrained Cys , usually referred as the backdoor cysteine residue . We believe that the formation of cyclic sulfenyl amide has a slower kinetic rate as compared to disulfide bonds formation in these cases . In this work we have also identified two PFAM families , Glutamine amidotransferase and Carbon-nitrogen hydrolase , that lack experimental evidence of cysteine oxidation but have a relevant Cys in the active site . [80 , 90] In this sense , it would be interesting to conduct experiments to analyze possible redox regulation of members of these families . Besides the previous interesting findings we searched in our list for proteins that are exposed to stress conditions . One example of these cases is the protein tyrosine phosphatases ptpB from Mycobacterium tuberculosis ( Mt ) . This protein has been reported to be involved in bacterial resistance to oxidative stress conditions found inside the macrophage , by modulating the activity of several cytosolic proteins . The role of ptpB is not completely clear , although one study points to the blocking of ERK1/2 and p38 IL-6 production pathways and Akt activation in the host cell [91] . On the other hand , the ptpA phosphatase of Mt , has the same fold but the cysteine was reported to be in a beta-sheet conformation near the forbidden zone , which could be a bias towards a more likely psi-dihedral angle . PtpA has been shown to dephosphorylate VPS33B , a component of the phagosome-lysosome fusion machinery [92] , and has also been reported to bind to a proton ATPase subunit preventing the acidification of the phagosome[93] . Both proteins are key elements of the mycobacteria nitrosative stress response , and thus both proteins must act in an oxidative environment where Cys oxidation would be favored . In this scenario , the presence of a key cysteine in the forbidden-psi conformation would protect ptpA/B from oxidative damage , through the formation of the cyclic sulfenyl amide . Interestingly , we found that ptpA could be regulated by cyclic sulfenyl amide formation although it has not been detected . On the other hand , ptpB has an extra domain called “lid domain” which acts as a gate to the active site of this enzyme , protecting it from oxidative stress . [94] Cysteines residues have a rich chemistry and are involved in a plethora of redox reactions . Initial oxidation to sulfenic acid has been shown to be dependent on the cysteine pKa [96–97] . It has been previously shown experimentally that in PTP1B the reactive cysteine is predominantly deprotonated at physiological pH [86] , something necessary for the phosphatase activity , but also makes the cysteine susceptible to fast oxidation . In agreement , our simulations show that the pKa decreases because the cysteinate is stabilized by the structural motif present in PTP1B ( Also in other proteins identified in our study ) . We found that several NH groups from the backbone are able to perform hydrogen bonds with the negative sulfur atom due to the constrained cysteine and the beta-loop-helix motif . However , we found that the forbidden psi angle is not sufficient to lower the pKa as in the model peptide its value is similar to the one of free cysteine in water . Proteins that have a reactive cysteine in their active site that has a low pKa are susceptible to inactivation by radical species like H2O2 . The first step in this oxidation is the formation of sulfenic acid . In this work we found that a constrained conformation helps , once the sulfenic acid is formed , to protect its irreversible oxidation by forming a cyclic sulfenyl amide . According to our results the reaction mechanism that converts the sulfenic acid to a cyclic sulfenyl amide occurs through a seemingly dissociative mechanism , with a relative small free energy barrier . There is also a key role of the solvent that needs to be treated explicitly . Our findings indicate that the role of the protein in catalyzing the reaction is not due to the presence of nearby residues but to promote a constrained conformation necessary for the reaction to occur , as similar activation barriers are obtained in the protein and in a model peptide in water . Experimentally , the reaction yielding the cyclic sulfenyl amide has been shown to occur in PTP1B as well as in several small model compounds . [32 , 33 , 95] In the protein , it is clear that the mechanism goes directly from the sulfenic acid to the cyclic product , and although its rate has not been measured , it should be able to compete with further Cys oxidation ( to sulfonic acid ) and thus protect the Cys in a physiological context . The relative small barrier obtained in the present work is consistent with this idea and underscores the likelihood of the presented mechanism . Last but not least , our results allow us to propose that the reaction should be promoted in acidic media , and thus show a pH dependent rate . Moreover , in the biological context , the presence of oxidative stress is usually accompanied by acidic conditions , and thus protection of key Cys through the present mechanism could be promoted . We have also identified two families of proteins that have the constrained cysteine in several of its members but lack the beta-loop-helix motif . According to our QM/MM results on a model peptide those proteins could form the cyclic sulfenyl amide if the cysteine is oxidized to sulfenic acid . However , in the Retroviral aspartyl protease family no oxidation of the constrained cysteine has been reported while in the Beta-lactamase2 family oxidation has been reported but only to disulphide bond . As we have previously proposed for the PTEN family , if a backdoor cysteine is present , disulphide formation seems to be preferred to sulfenyl amide . Despite the fact that we have identified 125 proteins that only have the constrained cysteine , there is no experimental evidence , mainly due to the lack of available structural information . Moreover , the formation of sulfenic acid , has not been reported in these proteins . Overall we have identified a group of proteins ( 270 ) that have a constrained Cysteine , located in a “forbidden” region of the Ramachandran plot ( psi angle: -150 to -90° ) , that according to our QM/MM results , enhances the formation of sulfenyl amide when the Cysteine is oxidized to sulfenic acid . We also describe a subset of proteins ( 145 ) that have a beta-loop-helix motif which allows them to lower the pKa enhancing their catalytic activity and also their reactivity towards ROS . In this subset , the constrained cysteine seems to be necessary for protection of the Cys residue from further oxidation as the cyclic sulfenyl amide can be then be recovered .
|
Cysteine oxidation is emerging as a relevant regulatory mechanism of enzymatic function in the cell . Many proteins are protected from over oxidation by reactive oxygen species by the formation of a cyclic sulfenyl amide . Understanding how cyclic sulfenyl amide is formed and its dependence on protein structure is not only a basic question but necessary to predict which proteins may auto protect from over oxidation We describe a structural motif , which includes cysteine residues with a constrained conformation in a “forbidden” region of the Ramachandran plot plus a Beta-Cys-loop-helix motif , which has a reactive low pKa Cysteine and also enables to form the cyclic sulfenyl amide with a low activation barrier . Our QM/MM computations show that the cyclization reaction only occurs if the “forbidden” conformation is acquired by the Cysteine residue . This structural motif was identified at least in 7 PFAM families and 145 proteins with solved structure , showing that a large number of proteins could have the ability to go through such cyclic product preventing irreversible oxidation .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"QM/MM",
"methods",
"Results",
"Discussion"
] |
[] |
2015
|
Protein Topology Determines Cysteine Oxidation Fate: The Case of Sulfenyl Amide Formation among Protein Families
|
Head and Neck Squamous Cell Carcinoma ( HNSCC ) is a heterogeneous disease of significant mortality and with limited treatment options . Recent genomic analysis of HNSCC tumors has identified several distinct molecular classes , of which the mesenchymal subtype is associated with Epithelial to Mesenchymal Transition ( EMT ) and shown to correlate with poor survival and drug resistance . Here , we utilize an integrated approach to characterize the molecular function of ETS1 , an oncogenic transcription factor specifically enriched in Mesenchymal tumors . To identify the global ETS1 cistrome , we have performed integrated analysis of RNA-Seq , ChIP-Seq and epigenomic datasets in SCC25 , a representative ETS1high mesenchymal HNSCC cell line . Our studies implicate ETS1 as a crucial regulator of broader oncogenic processes and specifically Mesenchymal phenotypes , such as EMT and cellular invasion . We found that ETS1 preferentially binds cancer specific regulator elements , in particular Super-Enhancers of key EMT genes , highlighting its role as a master regulator . Finally , we show evidence that ETS1 plays a crucial role in regulating the TGF-β pathway in Mesenchymal cell lines and in leading-edge cells in primary HNSCC tumors that are endowed with partial-EMT features . Collectively our study highlights ETS1 as a key regulator of TGF-β associated EMT and reveals new avenues for sub-type specific therapeutic intervention .
Head and neck squamous cell carcinoma ( HNSCC ) is the sixth most common cancer world-wide , with an annual incidence of roughly 600 , 000 cases; 50 , 000 cases are diagnosed annually in the United States alone[1 , 2] . HNSCC carries a 5-year mortality rate of nearly 50% , making it a leading cause of cancer-related death[3] . The major etiologies of this cancer are related to tobacco and alcohol consumption as well as infection with human papillomavirus ( HPV ) , which has emerged in the past two decades as a prominent risk factor leading to a substantial proportion of HNSCCs diagnosed each year[4–8] . Common treatment regimens for HNSCC include surgery , radiation and cytotoxic chemotherapy . However , in the past 15 years , survival has not improved significantly , especially in patients with advanced and recurrent disease , which highlights the need for better understanding of the disease and personalized therapies [6 , 9] . A challenging aspect of HNSCC biology is that it is inherently a diverse disease , comprised of tumors from distinct anatomical locations spanning the entire upper aerodigestive tract[8] . In addition , HNSCC presents with enormous variability in patient prognosis as well as response to therapy , which may denote an underlying molecular heterogeneity[10 , 11] . Indeed , recent genomic sequencing studies of HNSCC have revealed this complex heterogeneity , and importantly also highlighted the lack of consensus and actionable driver mutations[12–16] . Initial grouping of HNSCC patients based on the unsupervised clustering of global gene expression revealed four major subtypes that were designated as Basal , Mesenchymal , Classical and Atypical[17 , 18] . However , the generation of new datasets , particularly those enriched in epigenomic features and the application of advanced bioinformatics tools for meta-analysis have begun to offer us a more detailed view of the various HNSCC molecular subtypes[19–21] . Importantly , it is becoming clear that each subtype of HNSCC possess a distinct pattern of gene expression that sets it apart in terms of the underlying biology , key drivers and potentially actionable targets . Notwithstanding the emerging nuances of the clustering profiles of HNSCC , the Mesenchymal subtype of HNSCC remains a persistent entity across various datasets and warrants further investigation due to its aggressive behavior . As implied in its name , the mesenchymal subtype of tumors show enriched expression of key drivers of the Epithelial-Mesenchymal Transition ( EMT ) and those associated with functional pathways of cell motility and angiogenesis[22–24] . Of note , HNSCC tumors characterized by elevated levels of classical EMT markers exhibit high rates of distant metastasis and shorter metastasis-free survival[25 , 26] . The distinct dysregulated gene expression programs which control phenotypes of the various subtypes of HNSCC , including the ETS1high mesenchymal , are likely to be driven by specific transcriptional regulators[27] . However , our current knowledge of the identity of such factors and understanding of their mechanisms of action remain sparse . Here , we have utilized a comprehensive epigenomic and transcriptomic approach to uncover a novel regulator of the Mesenchymal subtype of HNSCC . We show that ETS1 , an oncogenic transcription factor ( TF ) and prototypic member of the ETS family of TFs , is preferentially expressed in Mesenchymal HNSCC across several large-scale transcriptomic datasets . We comprehensively probed the molecular function of ETS1 within a Mesenchymal HNSCC cell-line model , SCC25 , by first identifying its direct target genes using a combination of ChIP-Seq analysis , and RNA-Seq analysis . Our studies uncovered a core ETS1 driven Mesenchymal gene signature , which identifies select genes that are highly correlative with the EMT process . Significantly , we observed that one major role of ETS1 could be in driving and maintaining the Mesenchymal state of HNSCC through the TGF-β signaling cascade . Our findings have likely relevance for human HNSCC tumors as exemplified by an enriched ETS1 gene signature in subpopulations of cells exhibiting a partial-EMT state that have been uncovered by recent single cell RNA-Seq ( scRNA-Seq ) analysis[28] .
Although an oncogenic role of ETS1 in HNSCC has been previously reported , such studies have been limited in scope predominantly to immunostaining of tumors from a relatively small number of patients[29–31] . Therefore , here , we examined available transcriptomic datasets to re-assess the ETS1 expression profile in various stages of HNSCC [14 , 32 , 33] . Our analysis revealed that across multiple microarray ( Fig 1A and 1B ) and RNA-Seq ( Fig 1C ) based datasets , ETS1 mRNA levels were consistently high in tumors compared to normal tissues and interestingly , showed increasing expression as tumors progressed from dysplastic to carcinoma states . The tumor-enriched expression of ETS1 was most strikingly discernible in the RNA-Seq data from the Cancer Genome Atlas ( TCGA ) , where ETS1 levels were higher in ~80% of the patient tumor samples compared to matched normal tissue ( Fig 1C ) . The high levels of ETS1 was associated with poor prognosis as evident from Kaplan-Meier analysis of a cohort of 97 HNSCC patients , which showed that patients with high ETS1 expression had a statistically significant overall worse outcome than patients with low expression ( Fig 1D ) . While patients with low ETS1 expression had a median overall survival time of 61 months , the median overall survival time of patients with high levels of ETS1 expression was reduced to 34 months . Furthermore , ETS1 expression in these patients was shown to significantly correlate with higher clinical stage ( Fig 1E ) with higher levels specifically associated with the late stages ( III/IV ) of cancer . The molecular heterogeneity of HNSCC is clearly evident by the four major gene expression subtypes that have been derived from clustering analysis of both microarray and RNA-Seq datasets generated from tumor patients[14 , 17 , 32 , 33] . We sought to both confirm these findings as well as generate an improved subtype specific gene-signature by performing differential gene expression analysis ( DEG ) on the updated and much larger TCGA RNA-Seq dataset of 504 HNSCC tumors . Hierarchical clustering analysis based on the mRNA expression level of the most highly expressed subtype-specific genes reaffirmed the existence of the four defined subtypes—Atypical ( 26 . 2% ) , Basal ( 30 . 6% ) , Classical ( 17 . 6% ) , and Mesenchymal ( 25 . 6% ) ( Fig 2A ) . As expected , these subtypes were associated with biological functional pathways and processes characteristic of their distinct molecular and cellular attributes . Indeed , gene ontology enrichment analysis of the Mesenchymal ( MS ) subtype tumors highlighted several processes associated with EMT such as the remodeling of the ECM , cellular locomotion , cellular differentiation in addition to EMT itself ( Fig 2B , S1 Table , MSigDB Biological Processes ) [34 , 35] . In contrast , examination of the ontology enrichments for the other three intrinsic subtypes yielded results that matched closely to the original findings from the subtype analysis [19 , 36] ( S1 Fig , S1 Table ) . Interestingly , we observed a significant enrichment of ETS1 expression ( Fig 2C ) in the MS subtype whereas a closely related family member , ETS2 did not show any sub-type specific expression pattern ( S1 Fig ) . To examine how well ETS1 expression correlates with EMT , we next used a recently devised robust EMT scoring method [37]to assess and rank each MS subtype tumor based on their level of EMT . As illustrated in a heatmap highlighting the range of EMT and the expression levels of selected subset of key EMT-related markers , tumors with the highest EMT score were also associated with the highest levels of ETS1 expression ( S2A Fig ) . Notably , examination of the distribution of the EMT score across the 4 subtypes of HNSCC clearly showed the MS subtype to have much higher level of enrichment ( Fig 2D ) . To validate the findings from this analysis , we examined the respective protein levels of EMT-related genes in HNSCC tumors derived from the TCGA Reverse Phase Protein Array dataset ( RPPA ) [38] . We observed similar overall trends in the protein and mRNA transcript levels of the genes shared between the RPPA and RNA-Seq datasets , as well as strong correlation of ETS1 protein with EMT Score , in agreement with the ETS1 mRNA level ( S2B Fig ) . One challenge of working with the available TCGA datasets is the confounding effects of tumor purity on the correlating and clustering analysis of global gene expression . To address this limitation , we next examined RNA-Seq data from 38 HNSCC Patient Derived Xenografts ( PDX ) [39] . We reasoned that the PDX datasets are likely to offer a more tumor-centric analysis due to the fact that the human stromal component of PDX models is often lost in early passages[40] and that the mouse stromal contribution can be eliminated during downstream RNA-Seq processing . Interestingly , even with a relatively small sample size , gene expression profile ( Fig 2E ) in the HNSCC PDXs mirrored the established distribution of the HNSCC subtypes ( 21% Atypical , 26 . 3% Basal , 13 . 2% Classical and 39 . 5% Mesenchymal ) and importantly , ETS1 expression was preferentially enriched in Mesenchymal subtype ( Fig 2F ) similar to TCGA datasets . Taken together , our analysis across multiple platforms revealed that ETS1 expression is preferentially enriched in the Mesenchymal subtype of HNSCC and it is highly correlated with the EMT score of tumors . For follow up studies to our observations from tumor datasets and to identify a suitable cell line for downstream function analysis , we next performed RNA-Seq analysis of eight representative HNSCC cell lines . Unsupervised hierarchical clustering analysis of gene expression profiles of the HNSCC cells allowed us to categorize them based on the intrinsic subtype classification ( Fig 3A ) , with two of the tongue-derived SCC cell lines , SCC4 and SCC25 matching the MS subtype . The SCC25 cell line , in particular exhibited the highest level of expression of ETS1 mRNA and importantly , the most enriched EMT score ( S2C Fig ) . The transcriptomic data based findings were corroborated by Western blot analysis with anti-ETS1 antibodies which confirmed relatively higher levels of ETS1 protein in the SCC25 , compared to other cell lines ( Fig 3B ) . The observation that the HNSCC cell lines retain subtype-specific gene expression profiles prompted us to further validate this in additional publicly available datasets . Similar analysis of RNA-Seq data of HNSCC cells generated by the Cancer Cell Line Encyclopedia Project as well as those from a recently published study[41 , 42] yielded similar results , whereupon cell lines of a specific subtype clustered together and segregated into four distinct molecular groups ( S3 Fig ) . The SCC25 cell line again showed high levels of ETS1 and a EMT-high mesenchymal gene expression program in these two broad datasets and hence was chosen for downstream analysis of ETS1 function . To examine the molecular mechanisms of ETS1 function , we next identified its global DNA-binding sites in SCC25 cells by ChIP-Seq experiments using two different anti-ETS1 antibodies . By using the ENCODE recommended IDR ( Irreproducible Discovery Rate ) method for data analysis , we identified 26 , 760 ETS1 genomic-bound sites that showed a high correlation among the ChIP-Seq replicates of three independent experiments ( Fig 4A , S2 Table ) . As expected , de novo motif analysis using MEME revealed the core ETS DNA-Binding motif as the most enriched motif in the ETS1-bound sites . This was further verified by TOMTOM based motif comparison , which revealed that the ChIP-Seq derived motif was nearly identical to the ETS1 DNA-Binding Motif ( p = 7 . 36e-6 ) [43] ( Fig 4B ) . The distribution of the ETS1 peaks relative to transcription start sites ( TSS ) showed that ETS1 preferentially bound to distal regulatory regions with the majority of sites being classified as intragenic ( Fig 4C ) . We investigated the ETS1-genomic interactions in the chromatin context by analyzing the histone modification profile around each individual ETS1 binding site , using H3K27Ac , H3K27Me3 , H3K4Me1 , and H3K4Me3 ChIP-Seq data from the SCC25 cell line . This cohort of histone marks , in various combinations can be used to identify active enhancers ( H3K27AcHigh:H3K4Me1High ) , active promoters ( H3K27AcHigh:H3K4Me3High ) and poised enhancers ( H3K27Me3High:H3K4Me1High:H3K27AcLow ) [44] . By calculating the normalized levels of histone modification ChIP-Seq signal at each ETS1 bound genomic region and performing k-Means clustering of differentially enriched epigenetic profiles , we identified three clusters of ETS1 binding . ( Fig 4D , S3 Table ) . While most of the ETS1 binding overlapped with active regulatory elements characterized by the tell-tale histone modification features of enhancers and promoters , a third distinct cluster of ETS1 binding was found to exhibit epigenetic features of bivalent and poised chromatin domains . Notably , these 3 clusters of ETS1 bound genomic regions were associated with genes associated with different enriched biological processes ( Fig 4E ) . The top enriched pathways for genes in the cluster 1 was for the maintenance of stem and epidermal cell differentiation , whereas genes associated with cluster 2 were associated with house-keeping functions such as transcriptional initiation as well as chromatin organization . Interestingly , cluster 3 represented genes involved in the negative regulation of epithelial cell differentiation , as exemplified by the enrichment of genes involved in the inhibition of Notch signaling pathway . The interesting observation that ETS1 shows a contrasting bias of enriched binding to promoters for house-keeping genes and to distal enhancers that are likely to drive tissue-specific gene expression programs is consistent with previous ETS1 ChIP-seq results from other cell types[45] . Overall , this analysis suggested a global role of ETS1 as a transcriptional regulator of the cellular stem and differentiation state of SCC25 cells in particular , and by extension possibly other epithelial cancer cells in general . The dynamic interaction of ETS1 with distinct chromatin states prompted us to next examine Super-Enhancers ( SEs ) that are typically populated by master TFs and have the highest content of the active chromatin mark H3K27ac[46–48] . We generated the SE landscape of SCC25 by implementing the Rose algorithm with the H3K27Ac ChIP-Seq data and identified 965 such regulatory elements ( S4 Table ) . As expected , several SEs are associated with master regulators and critical components of epithelial development and differentiation , such as EGFR and TP63 , in agreement with the squamous origin of the SCC25 cell line ( Fig 5A , S4 Fig ) . Additional SE-marked genes included those related to stem cell biology and EMT , such as SOX9 and Vimentin and importantly , ETS1 itself ( Fig 5A , S4 Fig ) . Interestingly , in keeping with the functional importance of SEs , SE-associated genes exhibit higher median expression compared to those associated with Typical Enhancers ( TEs ) and show enrichment of signaling pathways and processes linked to oncogenic hallmarks such as angiogenesis and the p38 MAPK signaling , in addition to pathways associated with EMT , such as PDGF and WNT signaling ( Fig 5B and 5C ) . To identify potential TFs that might preferentially bind to SEs , we next performed motif enrichment analysis of nucleosome free regions ( NFRs ) within the SCC25 SEs . By using the motif enrichment tool AME and querying the JASPAR motif database [49] , we found that the ETS motif was one of the most-enriched motif based on multiple hypothesis adjusted p-value , along with AP-1 , a known co-factor[50] of ETS proteins ( Fig 5D ) . Indeed , compared to TEs , where only a quarter were bound by ETS1 , almost all SEs ( 98% ) were occupied by ETS1 , with dense ETS1 binding of an average of nearly 5 times per SE ( Fig 5E ) . This preferential association of ETS1 to SEs is further evident when the average ETS1 ChIP-Seq signal per NFR was compared between SEs and TEs ( Fig 5F ) . Taken together , these findings suggest preferred binding of ETS1 in the SEs of SCC25 as a mode to regulate various important cellular facets of HNSCC . In order to investigate the functional relevance of ETS1 and to identify its target genes , we next undertook knockdown ( kd ) studies . Towards this end , we used a lentiviral based system to generate SCC25 cells that stably express previously validated ETS1-targeting shRNAs [51] or a control , non-targeting shRNA ( shCON ) . Efficient and specific knockdown of ETS1 was achieved by two independent shRNAs , sh2 and sh3 , as evident by western blot analysis which showed robust loss of ETS1 expression but no effect on closely related ETS2 protein ( Fig 6A ) . Based on results obtained from several independent experiments , ETS1kd by sh3 was more robust and consistent , in agreement with published results [51] . Hence for most of the subsequent studies , SCC25 stable cells with sh3-mediated knockdown were utilized and independent verification of key findings were performed by ETS1kd by sh2 , wherever necessary . Reduced ETS1 expression and/or overexpression of ETS1 in other cell lines , including prostate and breast cancer cells have been previously shown to influence cell growth and migration[52–55] . To examine if ETS1 behaved similarly in SCC25 cells , we examined the effect of the loss of ETS1 using a MTT proliferation assay . A significant reduction in percent cell growth was observed in the cells expressing the ETS1kd compared to the control ( Fig 6B ) . To follow-up on these observations , we assessed the requirement of ETS1 for the SCC25 cells to migrate and invade in both 2D and 3D cell culture model systems . We first used the wound healing scratch assay to assess the ability of a confluent mono-layer of the stably transfected SCC25 cells to invade into identical denuded , cell-free areas within a cell culture plate . To negate the confounding effects of the cell proliferation defects upon loss of ETS1 , we performed the wounding experiment after treating the cells with mitomycin , a proven cell growth inhibitor . Loss of ETS1 greatly inhibited the ability of SCC25 cells to migrate into the cell-free zone since after 48 hours , while the control cells completely closed the wound , the ETS1-depleted cells managed to invade only ~20% ( sh3 ) or 40% ( sh2 ) of the area ( Fig 6C ) . Next , we tested the ability of ETS1-depleted SCC25 cells to invade through 3D matrix by using the Boyden chamber invasion assay . The control SCC25 cells had nearly 7 times more cells invade as compared to the SCC25-ETS1kd cells ( S5 Fig ) . To confirm these findings , we next decided to perform similar independent experiments with the SCC4 cell line , which is not only mesenchymal in its gene expression traits but also has high expression of ETS1 . Stable SCC4 cells were established for ETS1 knockdown by shRNAs , sh2 and sh3 as in SCC25 cell line . The effects of loss of ETS1 was more pronounced in SCC4 cell line such that the cells infected with the more potent sh3 were severely growth arrested and could not be propagated long term in culture easily . However , MTT proliferation and wound healing scratch assays with SCC4 control and ETS1sh2kd cells recapitulated the findings of SCC25 cells with loss of ETS1 leading to inhibition of cell growth and migratory capacity , respectively ( S6 Fig ) . Taken together , these results confirmed previous findings and supported the notion that ETS1 plays an important role in the growth , proliferation and migratory potential of the HNSCC mesenchymal cells . As a means to identify ETS1 targets , we next performed RNA-Seq experiments to profile the global transcriptomic changes that are unleashed upon loss of ETS1 . For these studies , we employed both siRNA transfected SCC25 cells and the previously described shRNA transduced SCC25 stable cell lines to examine the short and long term effects of ETS1 loss , respectively . As to be expected from prolonged loss of ETS1 , shRNA mediated ETSkd resulted in a significant alteration in the mRNA landscape of the SCC25 cells with 6727 genes showing statistically meaningful changes ( S5 Table ) . In contrast , the changes associated with the siRNA mediated knockdown were modest , resulting in 786 genes with significant up or down regulation despite the marked reduction in the levels of ETS1 transcripts in the ETS1-siRNA treated cells . Interestingly , there was a 67% overlap in Differentially Expressed Genes ( DEGs ) profile between the two experimental conditions ( S5 Table ) . This raised the possibility that some of the effects of loss of ETS1 are transitory in nature–indeed there were 243 genes that were altered only in the siRNA experiment . This cohort of DEGs that were exclusive to the siRNA-mediated knockdown of ETS1 were involved in PI3K signaling , as well as processes that have been associated with EMT in cancer such as inflammation and FGF signaling . A number of altered genes also included members of the WNT and Notch signaling cascades , which are involved in the maintenance of embryonic stem cell pluripotency and have also been linked to EMT and cancer stem cells ( Fig 7 ) . Since the long-term effects of constitutive loss of ETS1 are likely to be pronounced , we followed up on a larger cohort of genes that showed differential gene expression in response to the shRNA mediated knockdown of ETS1 . We found that a high percentage of the significantly altered genes ( 5991 out of 6727 ) were also direct targets of ETS1 based on ChIP-Seq data ( S5 Table ) . To facilitate further analysis , we focused on genes that change at least 2-fold in expression , and identified 1038 such transcriptional targets of ETS1 . We first used the Ingenuity Pathway Analysis ( IPA ) Regulator Effects tool to match significantly altered genes with their associated biological phenotypes . Interestingly , the analysis highlighted a number of predicted biological effects due to loss of ETS1 , such as tumor cell migration and invasion , cell cycle progression and tumor growth , ( S7 Fig ) . Next , we correlated the function of ETS1 to known upstream transcriptional regulators as well as the signaling cascades under their control using the IPA Upstream Regulator tool . Interestingly , we observed that the one major subset of ETS1-driven DEGs , consisting of 233 genes were significantly associated with the activation of TGF-β signaling ( Fig 7 ) . The link between TGF-β signaling and its role in inducing EMT has been established across numerous cancer models[56–58] . Our findings suggested that ETS1 could function as a driver of the mesenchymal phenotype of the SCC25 cells by acting as a regulator of TGF-β signaling . We further validated ETS1/TGF-β link by examining the enrichment of ETS1 direct targets in established gene-signatures related to TGF-β activation in cancer . We observed a significant overlap of 188 ETS1 target genes , with 118 of those displaying perfect correlation with the DEGs ( Fig 8A , S6 Table , p = 4 . 0e-28 , Fisher’s exact test ) when compared with a comprehensive EMT gene signature that has been recently developed[59] . Given the likelihood that ETS1 could be driving and maintaining the mesenchymal gene expression program of SCC25 through the regulation of the TGF-β signaling pathway , we next examined the activation status of Smad2 , the major receptor-activated Smad downstream of TGF-β signaling . For this purpose , the stably transfected control and the ETS1 sh2 and sh3 transduced SCC25 cells were stimulated with 1ng/ml and 5ng/ml of TGF-β1 for 1 , 3 and 24 hours and examined by western blot analysis . Interestingly , we observed a dramatic decrease in overall levels of p-Smad2 in the ETS1kd cells as compared to the control ( Fig 8B ) . Additionally , we observed a decrease in the overall phosphorylation levels of p-Smad2 over time , which was not observed in the control cells; the shCON cells maintained high levels of p-Smad2 after TGF-β1 stimulation in all durations of treatment . Similar effects of TGF-β signaling , specifically the reduced levels of p-Smad2 in the ETS1kd as compared to the control cells were also observed for SCC4 cells , particularly after 24 hours of TGF-β treatment ( S8A Fig ) . These findings suggest ETS1 plays a role in maintaining the overall activation state of the TGF-β signaling pathway . The strong association of ETS1 function with TGF-β induced EMT pathway prompted us to further validate this relationship with an independent gene-signature aimed at quantifying the EMT status of a tumor . Towards this end , we utilized a EMT signature that has been derived from the common molecular EMT features of several epithelial cancer types[60] and found a significant overlap with the SCC25 ETS1 target genes ( Fig 9A , S7 Table , 143/315 , p = 1 . 49e-9 Fisher’s Exact Test ) . As evident from the heatmap , a number of key EMT related genes , such as VIM , AXL and SNAIL2 were downregulated after knockdown of ETS1 . In contrast , genes representing the reverse state of Mesenchymal Epithelial Transition ( MET ) were upregulated , implying a potential repressive role for ETS1 in keeping the epithelial state of SCC25 suppressed ( Fig 9B ) . We independently confirmed the altered expression of the protein levels of some of these EMT/MET associated genes by western blot analysis of both the SCC25 and SCC4 control and ETS1sh2 and sh3 knockdown cells ( Fig 9C , S8B Fig ) . The strong association of ETS1 with the MS subtype of HNSCC tumors prompted us to compare the mesenchymal specific gene signature derived from the TCGA RNA-Seq dataset with the ETS1 target genes in SCC25 . There was a significant overlap between the two sets of gene lists ( p = 1 . 84e-5 , Fisher’s Exact Test ) resulting in the identification of a core ETS1 driven 169 Mesenchymal gene signature ( ETS1 MGS , S8 Table ) . Interestingly , nearly 18% of these genes were found to associated with a SE and included AXL , a known regulator of cancer associated EMT[61] . Spurred by these findings , we decided to test whether the ETS1 MGS could act as an independent means of classifying HNSCC , especially as the core ETS1 MGS did not include any of the classical EMT markers . After classifying each tumor as EMT-High , EMT-Mid and EMT-Low , based on the distribution of EMT Scores , we utilized principal component analysis ( PCA ) to measure and visualize the variance in gene expression within TCGA HNSCC Tumors as a function of the ETS1 MGS . The resulting PCA plot displayed a striking degree of variability between the different tumors , however; the centroids of each group of tumors appeared as clearly separated entities ( S9 Fig ) . Indeed , the first two principal components were associated with the total expression of the ETS1 MGS , which in turn was highly correlated with the distribution of EMT scores within the tumors . This finding indicated that the ETS1 MGS was able to cluster HNSCC based upon an intrinsic Mesenchymal gene expression program . As a final test of a possible functional role of ETS1 as a regulator of EMT in tumorigenesis , we probed the scRNA-Seq datasets of HNSCC that have been described recently [28] . In this study , a sub-population of cells were identified that were defined by a partial-EMT ( p-EMT ) gene-expression program , and enrichment for TGF-β signaling . To identify a possible association of ETS1 with the p-EMT signature , we examined the enriched genes identified in these cell populations to see if they are also represented in the ETS1 MGS . Interestingly , we observed a significant overlap between the ETS1 MS Signature derived from our study and the top 100 genes enriched in p-EMT cells ( p-value = 7 . 144e-12 , hypergeometric test ) as well as a significant overlap with the non-negative factorization analysis derived p-EMT gene signatures ( p = 2 . 514e-21 , hypergeometric test ) ( S10 Fig ) . Taken together these findings are highly suggestive of a potential role of ETS1 in maintaining the EMT state in HNSCC , possibly as a master regulator of the TGF-β signaling pathway .
The complex heterogeneity of HNSCC , similar to cancers from other anatomical locations , is clearly evident in the distinct gene expression profile of its subtypes . To better understand this heterogeneity , it is important to dissect the complex molecular circuitry that governs the dysregulated transcriptional networks in a given subtype . One powerful approach towards this end is to identify and characterize the master TFs which are likely to drive , maintain and elicit the tell-tale characteristics of a given tumor subtype . Among such oncogenic TFs , the ETS family of proteins are emerging as crucial mediators of tumorigenesis . Indeed , ETS factors are aberrantly activated in cancer by a slew of molecular mechanisms that include chromosomal translocation , overexpression , or post-translational modifications[62] . These ETS driven alterations in turn induce dysregulated gene expression programs that are likely to be essential in various facets of tumorigenesis . As a prototypic and founding member of the ETS family , ETS1 has been widely studied and implicated in a variety of solid tumors[63] . Overall , data from such studies have often linked ETS1 expression to advanced state of epithelial cancers with poorer differentiation , higher invasive activity and angiogenesis , an increased risk to metastasis and a higher tendency to acquire drug resistance–all hallmark features of cancer . Despite such clear associations , the molecular attributes of ETS1 , in particular its target gene repertoire and function in context of a defined tumor subtype remains ill-understood . Here we have comprehensively examined ETS1 in HNSCC , where we find ETS1 to be specifically enriched in expression across cell lines , human tumors and PDXs representing the mesenchymal subtype and associated with poor tumor outcome . To gain mechanistic insights into its function , we have undertaken genomic and epigenomic based identification and analysis of the ETS1-dependent cistrome in the SCC25 cell line , which has revealed several interesting findings . Foremost , we uncovered that the direct and indirect target genes of ETS1 are quite extensive and impinge upon virtually all key aspects of tumorigenesis . Our findings are in some ways different from similar studies performed in prostate cancer cells , [53] where the number of ETS1 bound elements are significantly smaller compared to those in the SCC25 cells . We suspect that this might be in part due to the robust nature of our ChIP experiments , having utilized two independent antibodies . Our ETS1-genomic interactions are further strengthened by the overlay of active histone marks , which has allowed us to gain exciting insights into the ETS1-chromatin interactions and the distinct clusters of regulatory elements that are bound by ETS1 . Importantly , the proclivity of ETS1 to be enriched for binding in the Super-Enhancer regions and for itself to be regulated by a SCC25 Super-Enhancer is broadly suggestive of its crucial function and specifically the existence of a potential feed-forward loop . It is worth noting that similar enrichment of ETS motifs and the binding of ETS2 have been reported in Super-Enhancers identified in other contexts , such as skin SCC and nasopharyngeal cancer[64 , 65] . The remarkable structural and functional similarity of ETS1 and ETS2 hints at an interesting crosstalk and possible redundant interplay between these closely related TFs in HNSCC subtypes . Recent studies however suggest that ETS1 and ETS2 can also direct gene expression programs in opposite directions allowing these TFs to switch between oncogenic and tumor suppressive functions in cell-type specific manner [66]–whether similar mechanisms are at play in HNSCC begs further investigation . We have performed extensive examination of the ETS1 target genes by integrating RNA-Seq and ChIP-Seq experiments . These studies have reaffirmed the notion that ETS1 casts a wide net in regulating the transcriptional network of cancer cells , affecting virtually all key hallmarks of cancer . Prominent among these ETS1-dependent processes and pathways are cell proliferation and migration , angiogenesis , inflammation and in keeping with the mesenchymal nature of the SCC25 cells , EMT . Interestingly , many of these processes are also affected in our previously described transgenic mouse model of ETS1 overexpression in squamous epithelium that leads to dysplasia and inflammation[67 , 68] . In this context , it is also worth noting that although ETS1 has not firmly secured its place alongside the classical EMT-TFs , prevailing evidence from past literature and our present study illustrate its growing and pervasive role in Type I ( developmental ) , Type II ( inflammation and wounding ) and Type III ( neoplasia ) EMT[63] . Indeed , a recent pan cancer survey of EMT markers across the TCGA has revealed ETS1 to be one of the highest ranked TF which showed significant EMT correlations within multiple tumor types that span various histological or cellular backgrounds[69] . We posit that our findings on the ETS1-TGF-β-EMT axis in SCC25 and SCC4 cells are considerably strengthened by similar trends in cells with partial-EMT that are localized to the leading edge of primary HNSCC tumors[28] . These associations might have implications for a range of other solid tumors , some of which are also endowed with mesenchymal features , such as Claudin-low breast cancer . Our findings are also supported by results from studies on CD90 ( + ) tumor initiating cell population from esophageal squamous cell carcinoma , where the deregulation of an ETS1/MMP signaling pathway and EMT figure prominently[70] . Taken together our discovery of ETS1 as a biomarker of mesenchymal HNSCC and a master regulator of key EMT genes offers promising new avenues for targeted therapeutic possibilities that are tumor subtype specific . This includes possible re-purposing of clinically approved drugs such as Dasatinib , which has been shown to target ETS1 for proteasomal degradation[51] .
SCC25 and SCC47 cell lines were purchased from ATCC and MilliporeSigma , respectively . Cell lines UMSCC29 , UMSCC23 and UMSCC103 were obtained from Dr . Thomas Carey ( University of Michigan ) . HSC-3 and CAL-27 cell lines were generously donated by Dr . Manish Bais ( Boston University ) . SCC4 cell line was obtained from Dr . James Rheinwald ( Harvard University ) . SCC25 and SCC4 was maintained in DMEM/F12 high glutamine culture media supplemented with 10% fetal bovine serum ( FBS ) , 400 ng/mL hydrocortisone and antibiotics . All other cell lines were maintained in high glutamine DMEM containing 10% FBS , 1% GlutaMax , 1% Non-Essential Amino Acids , and antibiotics . The identities of all cell lines were confirmed via STR profiling . For siRNA mediated knockdown of ETS1 , siGENOME siRNA , which consists of 4 pooled siRNAs , were used . Biological replicates of SCC25 cells were treated with either 75 pmol of non-targeting siRNAs or siRNAs targeting ETS1 , using Lipofectamine RNAiMAX . Lentiviral transduction of SCC25 and SCC4 cells expressing either a non-targeting shRNA or ETS1-targeting shRNAs ( shETS1-2 , shETS1-3 ) was achieved by the One-Shot LentiX Kit ( Clontech ) followed by puromycin selection ( 2 μg/mL ) . Cells were grown to ~90% confluency and proteins were extracted via addition of Laemmli Sample Buffer directly to the dish . Western blot analysis was performed using a standard protocol as described before ( 68 ) . The antibodies used were: ( C4 , Santa Cruz Biotechnology ) , p-Smad2 ( GTX111075 , GeneTex ) , Smad2 ( GTX111075 , GeneTex ) , Vimentin ( Epitomics ) , Snai2 ( Santa Cruz Biotechnology ) , ETS2 ( GTX104527 , GeneTex ) , N-Cadherin/CDH2 ( BD Biosciences ) , TP63 ( 4A4 , Santa Cruz Biotechnology ) , AXL ( Cell Signaling Technologies ) . The MAB374 antibody ( EMD Millipore ) was used to detected GAPDH as a loading control at 1:60 , 000 dilution . The control shRNA and ETS1 shRNA transduced SCC25 and SCC4 cells were seeded into a 96-well plate ( 104 cells/well ) and incubated for 24 , 48 , 72 and 120 hours , and then were subjected to the Vybrant MTT Cell Proliferation Assay Kit ( Thermo Fisher ) according to the manufacturer’s instructions . The viability of the cells was assessed by measuring the absorbance at 492 nm using a Cytation 5 Cell Imaging Multi-Mode Reader ( BioTek Instruments ) . All experiments were performed in triplicate . The cell proliferation differences were reported in percent change in growth , which measures the changes in cell number relative to the starting 24-hour time point . Both control and ETS1 knockdown of SCC25 and SCC4 cell lines were seeded into 6-well plates and grown until confluence . Cells were treated with 10 μg/mL of mitomycin C ( Enzo Life Sciences ) for 2 hours prior to the initiation of the wounding . A scratch was then made across the center of each well using a sterile pipette tip , and non-adherent cells were washed off with Phosphate-Buffered Saline . Cells were grown in regular media and images of the wound were taken immediately and after in 12–24 hour intervals until wounds were completely closed . All experiments were performed in triplicate and the data is presented as percent change in area relative to time 0 . The Corning Matrigel Invasion Assay ( Corning BioCoat Matrigel Invasion Chambers , 8 μm , Corning ) was used to characterize the ability of SCC25 cells to invade through a 3D matrix . Briefly , inserts were thawed at room temperature and reconstituted using serum-free DMEM/F12 media . SCC25 cells with non-targeting control or ETS1 targeting shRNA , were added to quadruplicate inserts at a concentration of 5x105 cells/mL in serum free media . Serum containing media was placed in the well beneath the insert and the entire insert-containing 24 well plate was allowed to incubate at 37°C at 5% CO2 for 24 hours . After incubation , the top of the membrane of the inserts was swabbed to remove cells that had not invaded through the matrix and membrane . The wells were then washed and aspirated and subsequently fixed in 100% Methanol for 10 minutes at -20°C . Methanol was removed and inserts were washed 3 times and subsequently stained using a DAPI stain ( Cell Biolabs ) . The insert membranes were placed on glass slides and visualized using the Cytation 5 Cell Imaging Multi-Mode Reader using the 405nm excitation laser and using the DAPI filter . Stained nuclei were counted using the supplied software in five separate fields of each insert . SCC25 and SCC4 cells stably expressing a non-targeting shRNA or ETS1 targeting shRNA , were plated into 6 well plates at 250 , 000 cells per well in duplicates . The cells were allowed to adhere to the plates for 12 hours and were subsequently serum starved for 24 hours . After serum starvation , the cells were either treated with a solution of 1ng/mL , 2ng/mL or 5 ng/mL TGF-β1 ( R&D Systems ) for a duration of 1 , 3 , or 24 hours . Post-treatment , whole cell protein lysates were collected and resuspended in Laemmli Sample Buffer ( Bio-Rad ) for western blot analysis . The High Sensitivity ChIP-IT Kit from Active-Motif and the associated protocol was used to perform ChIP . SCC25 cells were grown to ~90% confluency and cross-linked in the supplied fixation Buffer supplemented with 0 . 5% Formaldehyde for 10 minutes . Lysates from the fixed cells were subsequently sonicated with a Diagenode Bioruptor to obtain sheared chromatin with an approximate fragment length of 150–400 bp . The ChIPs for ETS1 were carried out using 4 μg of the polyclonal ant-ETS1 antibodies , C20x ( Santa-Cruz ) and A300-501A Antibody ( Bethyl Laboratories ) . After cross-link reversal , proteinase-K/RNAase-A treatment and DNA-purification , libraries were prepared using the ThruPLEX DNA-Seq Kit ( Rubicon Genomics ) . ChIP DNA and input controls were then subjected to 50 bp single-end sequencing on an Illumina HiSeq 2500 , which resulted in 15–25 million reads per sample . Microarray profiling of patient samples was obtained from GEO ( GSE45153 , GSE6791 , and GSE41613 ) . Each dataset was processed according to the protocol defined by the original study . The Cancer Genome Atlas ( TCGA ) RNA-seq expression datasets was downloaded from GEO ( GSE62944 ) . Briefly , featureCounts ( Rsubread v1 . 14 . 2 , [71] ) generated gene-level outputs from the entire TCGA RNA-Seq repository was utilized for downstream analysis . HNSCC samples ( 459 tumor and 44 normal ) were extracted and normalized using the median-ration method ( DESeq2 v1 . 12 . 3 , [72] ) and subsequently transformed to Transcript per Million ( TPM ) values[73] . Tumors were classified into intrinsic molecular subtypes as previously described[36] . EMT scores was calculated according to the formula defined by Salt et al . , 2014 . TPM tables were generated for each tumor within the TCGA HNSC RNA-Seq dataset and subsequently used as input for classification into one of four intrinsic subtypes as previously described [36] . FeatureCounts generated gene-level counts tables for each tumor was used for DESeq2 analysis whereby tumors of identical subtype classification were treated as replicates . All possible combinations of contrast was performed and a gene was considered subtype-specific if it was commonly over-expressed across each of the subtype-subtype comparisons . Total RNA from HNSCC cell lines was extracted using the Direct-zol RNA MiniPrep kit ( Zymo Research ) . For each RNA sample , cDNA libraries were prepared using the TrueSeq RNA Sample Preparation Kit ( Illumina ) , which were then 50 bp single-end sequenced on an Illumina HiSeq 2500 . Quality control metrics were performed on raw sequencing reads using the FASTQC v0 . 4 . 3 application . Reads were mapped to the Homo Sapiens genome ( GRCH37 build ) with TopHat2 v2 . 1 . 1 , using Bowtie2 v2 . 2 . 6 as the underlying aligner [74 , 75] . Reads aligning to the Ensembl GRCH37 build were quantified with featureCounts . Relative transcript abundances per experiment were reported in Transcript per million ( TPM ) values and were calculated from the featureCounts output generated by Rahman et al . [76] . All differential gene expression analysis was carried out using DESeq2 , using an FDR cutoff of 0 . 1 . Only protein coding genes with expression level of at least 1 TPM in a given replicate for each sample , were considered for DESeq2 analysis . Overall survival analysis of patients in the FHCRC dataset ( n = 97 , Lohavanichbutr et al . , 2013 ) was carried out using the survival R package ( v2 . 41–3 ) and Kaplan-Meier plots were generated using ggplot2 v2 . 2 . 1 . Patient tumor samples were deemed to have high ETS1 expression based on the median expression values . The log-rank statistic was used to calculate statistical significance between different patient classifications and their relationship to clinical outcome . The raw ChIP-Sequencing reads from all experiments were mapped to the Homo sapiens genome ( hg19 build ) using Bowtie v1 . 1 . 1 with the parameter , m = 1 , which removes all reads mapping to multiple genomic loci . All ChIP-Seq experiments were underwent strand cross-correlation analysis using SPP v1 . 10 . 1 and deemed acceptable if the QualityTag score was 1 or higher[77] . All identified peaks were matched to the nearest gene using GREAT analysis using default settings[78] . The MEME tool was used to generate a de-Novo motif from the ETS1 ChIP-Seq experiments [83 , 84] . A 600 bp window was selected around each ETS1 peak summit to use as an input for the program . The selected MEME-generate motif was subjected to TOMTOM motif comparison analysis , which showed that the motif derived from genomic segments bound by ETS1 as ascertained from ChIP-seq data was most similar to the consensus ETS core motif ( 43 ) . The comparison with the ETS1 motif is highlighted . The CEAS tools was used to annotate ETS1 ChIP-Seq peaks to the nearest genomic of feature of the hg19 genome assembly[85] . The promoter region was considered up to 500 bp away from a TSS and the proximal enhancer was considered 500 bp-1500 bp away . Any binding within a gene was considered intragenic , whereas any binding site greater than 1500 bp upstream or downstream was considered distal intergenic . Heatmap showing the ChIP-Seq signal of the H3K27Ac , H3K27Me3 , H3K4Me1 , and H3K4Me3 Histone modifications around a 2-kb window centered at each ETS1 ChIP-Seq peak summit . The resulting histone signal enrichment was subjected to k-Means clustering ( k = 4 ) . The fluff python package was used to generate the heatmap , using default parameters[86] . Detailed experimental methods of the epigenomic studies in the SCC25 are described in a separate manuscript currently under review ( Tsompana et al . , ) . Biological replicates of H3K27Ac ChIP-Sequencing data from SCC25 cells ( manuscript under review ( Tsompana et al , . ) were aligned to the human genome as described above . Narrow-Peaks were called using MACS2 v2 . 1 . 0 using the following parameters ( -p 0 . 01 –nomodel–extsize 150 ) . The resulting narrowPeaks files were converted to gff format and used as inputs for the ROSE Algorithm , which was run using default parameters along with appropriate input controls to generate typical and Super-Enhancers [47 , 48] . SCC25 H3K27Ac ChIP-Seq replicate experiments were merged . The deeptools program was used to convert the H3K27Ac signal to RPKMs ( Reads per Kilobase per Million Mapped Reads ) and was subsequently summed across the genome within 10 base-pair bins at the selected loci . To determine the top enriched DNA-Binding Motifs of transcription factors found within the Nucleosome Free Regions ( NFR ) of SCC25 Super-Enhancers , first NFRs were determined using the Homer findPeaks tool with the–nfr flag turned on . The AME tool was used to determine enriched motifs found within the JASPAR Vertebrates 2014 database . Motifs were ranked according to adjusted p-value . For ETS1 ChIP-Seq data , GREAT tool[87] was used to annotate binding loci to the nearest gene as well to carry out gene-set enrichment analysis using default parameters . For RNA-Seq data , the MSigDB was used to identify gene-sets with significant overlaps with the various gene signatures derived above using the hypergeometric test as well as controlling for multiple testing using the Benjamini-Hochberg correction ( cutoff = 0 . 05 ) . To visualize the ETS1 driven molecular signature and its relationship to possible biological phenotypes , Ingenuity Pathway analysis ( IPA , Qiagen ) was used . A causal network , which construct relationships between the changes in gene expression after knockdown of ETS1 and the downstream phenotypic and molecular changes occurring in SCC25 , was created using the Regulator Effect tool under default parameters . The expression of the ETS1 Mesenchymal Signature within the TCGA HNSC Dataset was used as input for PCA analysis . Each tumor’s EMT score was calculated and Fisher-Transformed . Tumors were classified as EMT High , Neutral or Low , based on the distribution of EMT scores , with Low being defined in the first quartile , High as the third and Neutral as in between .
|
The expression of the transcriptional regulator , E26 transformation-specific 1 ( ETS1 ) , is elevated in many epithelial cancers and portends aggressive tumor behavior and poor survival . Within these carcinomas , ETS1 function has been shown to be associated with a wide range of cellular responses that include increased proliferation , angiogenesis , metastasis and drug resistance . Here we focus on Head and Neck Squamous Cell Carcinoma ( HNSCC ) and discover that higher expression of ETS1 is specifically more pronounced in the mesenchymal subtypes of HNSCC , which represent tumors with enriched expression of Epithelial to Mesenchymal Transition ( EMT ) markers and inflammation . By using genomic and epigenomic strategies , we have identified the global targets of ETS1 in a preclinical Mesenchymal HNSCC cell model and determined the crucial gene network that is most dependent upon its function . We further validate this ETS1-driven gene expression signature within several HNSCC patient derived datasets and conclude that ETS1 acts as a crucial regulator of the TGFβ signaling cascade to drive EMT . Our findings reinforce the challenges of epithelial tumor heterogeneity and offer insights into molecular underpinning of a specific subtype that can be mined for cancer vulnerability .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"gene",
"regulation",
"carcinomas",
"cancers",
"and",
"neoplasms",
"oncology",
"regulator",
"genes",
"head",
"and",
"neck",
"tumors",
"gene",
"types",
"head",
"and",
"neck",
"squamous",
"cell",
"carcinoma",
"small",
"interfering",
"rnas",
"genomic",
"signal",
"processing",
"gene",
"expression",
"head",
"and",
"neck",
"cancers",
"biochemistry",
"signal",
"transduction",
"rna",
"carcinogenesis",
"cell",
"biology",
"nucleic",
"acids",
"genetics",
"tgf-beta",
"signaling",
"cascade",
"biology",
"and",
"life",
"sciences",
"squamous",
"cell",
"carcinomas",
"non-coding",
"rna",
"cell",
"signaling",
"signaling",
"cascades"
] |
2019
|
Molecular dissection of the oncogenic role of ETS1 in the mesenchymal subtypes of head and neck squamous cell carcinoma
|
Chronic schistosomiasis is associated with T cell hypo-responsiveness and immunoregulatory mechanisms , including induction of regulatory T cells ( Tregs ) . However , little is known about Treg functional capacity during human Schistosoma haematobium infection . CD4+CD25hiFOXP3+ cells were characterized by flow cytometry and their function assessed by analysing total and Treg-depleted PBMC responses to schistosomal adult worm antigen ( AWA ) , soluable egg antigen ( SEA ) and Bacillus Calmette-Guérin ( BCG ) in S . haematobium-infected Gabonese children before and 6 weeks after anthelmintic treatment . Cytokines responses ( IFN-γ , IL-5 , IL-10 , IL-13 , IL-17 and TNF ) were integrated using Principal Component Analysis ( PCA ) . Proliferation was measured by CFSE . S . haematobium infection was associated with increased Treg frequencies , which decreased post-treatment . Cytokine responses clustered into two principal components reflecting regulatory and Th2-polarized ( PC1 ) and pro-inflammatory and Th1-polarized ( PC2 ) cytokine responses; both components increased post-treatment . Treg depletion resulted in increased PC1 and PC2 at both time-points . Proliferation on the other hand , showed no significant difference from pre- to post-treatment . Treg depletion resulted mostly in increased proliferative responses at the pre-treatment time-point only . Schistosoma-associated CD4+CD25hiFOXP3+Tregs exert a suppressive effect on both proliferation and cytokine production . Although Treg frequency decreases after praziquantel treatment , their suppressive capacity remains unaltered when considering cytokine production whereas their influence on proliferation weakens with treatment .
The immune system has evolved several regulatory mechanisms to maintain immune homeostasis , prevent autoimmunity and restrain inflammation [1–3] . Many pathogens have developed mechanisms to manipulate the regulatory network of the host to their advantage , thereby generating conditions that ensure their survival for a prolonged period of time . In particular FOXP3+ regulatory T ( Treg ) cells have been shown to play a major role in the control of various parasitic infections suppressing local tissue damage and pathology that would result from otherwise over-reactivity . However , enhanced Treg cell activity may also allow the long-term survival of the parasite as the host is hampered from fighting the intruding pathogen effectively [4] . Schistosomiasis is a helminth infection affecting over 240 million people worldwide , especially children [5] . When chronic in nature it has been shown to be associated with general T cell hypo-responsiveness—evident from down-modulated antigen-specific Th1 and Th2 cell responses [6 , 7]] . This might result from mechanisms involving peripheral anergy and suppression triggered by regulatory cells , such as Tregs [8] . For example , in experimental murine models , it was observed that the presence of regulatory T cells was associated with suppressed development of pathology [9] , and down-modulated Th1 and Th2 responses [10 , 11] , promoting parasite survival within the host [12 , 13] . Evidence for Treg activity in human helminth infections has been provided by the detection of T cells with a regulatory phenotype in patients with lymphatic filariasis [14] , onchocerciasis [15 , 16] and schistosomiasis [17 , 18] . Effective chemotherapy with praziquantel has been shown to result in elevated antigen-specific proliferation and cytokine production , in particular interleukin ( IL ) -4 , IL-5 , and interferon ( IFN ) -γ [6 , 19–22] . Although , the frequency of Tregs , defined phenotypically as CD4+CD25hi without considering FOXP3 as a marker , decreased substantially after treatment with praziquantel [18] , their functional activity has not been studied before . The aim of this study was to assess whether S . haematobium infection in Gabonese children was associated with induction of regulatory T cells and to evaluate Treg activity during infection . To this end immune responses were evaluated before and 6 weeks after praziquantel treatment . Moreover , the functional activity of Tregs was assessed by comparing responses before and after their removal from peripheral blood mononuclear cells by in vitro magnetic depletion .
Ethical approval for the study was obtained from the Comité d’Ethique Régional Indépendant de Lambaréné . A signed informed consent form was obtained from parents or legal guardians of all children participating in the study . The longitudinal study was conducted at Centre de Recherches Médicales de Lambaréné ( CERMEL; formerly Medical Research Unit ( MRU ) of the Albert Schweitzer hospital ) . Participants were schoolchildren from a rural area ( PK15 ) highly endemic for S . haematobium approximately 15 km south of Lambaréné in the province of Moyen-Ogooué , Gabon [22 , 23] . Children were excluded from the study if they received anthelminthic treatment within the previous six weeks or had fever or other symptoms of acute illness . Fifty-two schoolchildren were recruited to participate in the study and of those twenty-eight ( mean age: 10 . 3y ( SD: 2 . 2 ) ; sex ratio: 15f/13m; median egg count/10ml urine: 72 . 5 ( IQR: 24 . 5–296 . 3 ) pre-treatment and 0 ( IQR: 0–1 . 5 ) at post-treatment; haemoglobin level: 11 . 1 g/dL ( SD: 1 . 0 ) pre-treatment and 11 . 1 g/dL ( SD: 1 . 0 ) post-treatment; and white blood cell level: 8 . 8 x103/mm3 ( SD: 3 . 2 ) pre-treatment and 9 . 4 x103/mm3 ( SD: 2 . 9 ) ) were included in the cellular analyses . Of the 24 schoolchildren that were not included in the final analysis , 4 did not return for post-treatment visit , 5 had less than 90% clearance of Schistosoma eggs at 2nd treatment and in 8 donors at pre-treatment and 11 donors at post-treatment Treg depletion failed . There was no significant difference between the schoolchildren that were included in the final analysis and those that were not . To compare the frequency of CD25hiFOXP3+ Treg cells between S . haematobium infected and uninfected schoolchildren an additional 10 S . haematobium-infected participants ( mean age: 12 . 5y ( SD: 1 . 5 ) ; sex ratio: 9f/1m; median egg count/10ml urine: 19 . 5 ( IQR: 3 . 3–216 . 5 ) ; haemoglobin level: 11 . 4 g/dL ( SD: 0 . 31 ) ; and white blood cell level: 8 . 8 x103/mm3 ( SD: 0 . 6 ) ) were recruited from the same rural area and seven uninfected subjects ( mean age: 12 . 9y ( SD: 2 . 6 ) ; sex ratio: 4f/3m; median egg count/10ml urine: 0; haemoglobin level: 10 . 9 g/dL ( SD: 0 . 7 ) ; and white blood cell level: 6 . 3 x103/mm3 ( SD: 0 . 7 ) ) were recruited from semi-urban Lambaréné . In addition to their positivity for S . haematobium infection , rural children also had significantly higher white blood cell levels ( p<0 . 05 ) compared to children from semi-urban Lambaréné . When comparing the 2 cohorts , while the additional cohort of children was a little older ( 12 . 7y vs 10 . 3y; p<0 . 05 ) , there was no significant difference in haemoglobin and white blood cell levels nor in infection intensity between the 2 groups . A midstream urine sample was collected during the day and 10 ml were passed through a 12 . 0 μm polyamide N filter ( Millipore ) for the detection of S . haematobium eggs by microscopy . Children were classified as infected if at least one S . haematobium egg was detected in the urine . An initial treatment with praziquantel ( 40 mg/kg ) was administered to Schistosoma-infected children at inclusion , and repeated after three weeks , in order to ensure clearance of parasites . Six weeks after initial treatment the efficacy of praziquantel was assessed by measuring egg load in urine . Donors were excluded from analysis , if their reduction in egg load was less than 90% following second treatment . All subjects who were egg-positive after the second treatment were given a third dose of treatment . Peripheral blood mononuclear cells ( PBMCs ) were purified from heparinized venous blood ( 7-10ml ) by Ficoll-Hypaque centrifugation ( Amersham Biosciences , Netherlands ) . Depletion of CD25hi T cells was performed using a suboptimal concentration of CD25 microbeads ( Miltenyi Biotec , Bergisch Gladbach , Germany ) according to the manufacturer’s instructions . This method has been shown by us in other studies to successfully deplete FOXP3 Tregs [24] . Similar results were obtained in Gabon as shown in S1A Fig . To analyse proliferation green-fluorescent dye carboxyfluorescein succinimidyl ester ( CFSE; Sigma-Aldrich , CA , USA ) was used; CFSE divides over daughter cells upon cell division and can be tracked by decreasing fluorescence intensity . CD25hi-depleted and total PBMCs were stained with 2μM CFSE for 15 minutes at room temperature prior to culture . After labelling , cells were cultured in RPMI 1640 ( Gibco , Invitrogen® , Carlsbad , CA , USA ) , supplemented with 10% fetal bovine serum ( FBS; Greiner Bio-One GmbH , Frickenhausen , Germany ) , 100 U/ml penicillin ( Astellas , Tokyo , Japan ) , 10 μg/ml streptomycin , 1 mM pyruvate and 2 mM L-glutamine ( all from Sigma-Aldrich , CA , USA ) . Cells were stimulated in 96-well round bottom plates ( Nunc , Roskilde , Denmark ) with medium , 10 μg/ml AWA , 10 μg/ml SEA , or 10 μg/ml BCG ( Bacille Calmette-Guérin; SSI , Copenhagen , Denmark ) and incubated in 5% CO2 at 37 . 5°C . After four days , supernatants were collected and stored at -80°C , while cells were harvested , fixed with 2% formaldehyde ( Sigma-Aldrich , CA , USA ) and , subsequently , frozen in RPMI 1640 medium supplemented with 20% FCS and 10% dimethyl sulfoxide ( DMSO; Merck KGaA , Darmstadt , Germany ) and stored at -80°C . After thawing , CFSE-labelled cells were incubated with CD4-PE ( SK3; BD Bioscience , San Diego , CA , USA ) and CD25-APC ( M-A251; BD Bioscience , San Diego , CA , USA ) , acquired on a FACSCalibur flow cytometer ( BD Biosciences , San Diego , CA , USA ) and data were analysed in a FlowJo Proliferation application ( Tree Star Inc . , Ashland , OR , US ) by calculation of the fraction of CD4+CD25hi cells that had divided from the starting population ( division index ) . To assess Treg depletion , ex-vivo PBMC were fixed with the FOXP3 fixation/permeabilization kit ( eBisocience , San Diego , CA , USA ) and frozen in RPMI 1640 medium supplemented with 20% FBS and 10% DMSO and stored at -80°C . For immunophenotyping isolated PBMCs were stained with CD4-PE ( SK3; BD Bioscience , San Diego , CA , USA ) , CD25-APC ( M-A251; BD Bioscience , San Diego , CA , USA ) and FOXP3-PE ( PCH101; eBioscience , San Diego , CA , USA ) , acquired on a FACSCalibur flow cytometer ( BD Biosciences , San Diego , CA , USA ) and data were analysed in a FlowJo software ( Tree Star Inc . , Ashland , OR , US ) . To assess the frequency of CD25hiFOXP3+ Tregs , ex-vivo PBMC were fixed with the FOXP3 fixation/permeabilization kit ( eBisocience , San Diego , CA , USA ) and frozen in RPMI 1640 medium supplemented with 20% FBS and 10% DMSO and stored at -80°C . For immunophenotyping isolated PBMCs were stained with CD4-PE/Cy7 ( SK3; BD Biosciences , San Diego , CA , USA ) , CD25-PE ( 2A3; BD Biosciences , San Diego , CA , USA ) and FOXP3-APC ( PCH101; eBioscience , San Diego , CA , USA ) , cells were acquired on FACSCanto II flow cytometer ( BD Biosciences , San Diego , CA , USA ) and analysed in FlowJo software ( Tree Star Inc . , Ashland , OR , US ) using Boolean combination gates . Cytokines were measured from supernatants using Luminex 100 IS System ( Invitrogen , Carlsbad , CA , USA ) and commercially available beads and standards from BioSource ( Bleiswijk , Netherlands ) for interferon-gamma ( IFN-γ ) , interleukin-5 ( IL-5 ) , IL-10 , IL-13 and IL-17 and tumor necrosis factor ( TNF ) . Beads were titrated for optimal dilution and used according to manufacturer’s instructions . Data analysis was performed using IBM SPSS Statistics version 20 for Windows ( IBM Corp . , Armonk , USA ) . Differences between groups were determined by the Fisher’s exact test for sex , by Mann-Whitney U test for S . haematobium infection intensity and by the independent student’s T test for age and haematological parameters . Cytokine concentrations in response to stimulation were corrected for spontaneous cytokine production by subtracting responses of unstimulated medium wells to obtain net cytokine responses , with negatives values set to half of the lowest value detected per given cytokine . To avoid type I and type II errors in multiple testing , immunological parameters were reduced by principal-components analysis ( PCA ) . First , R v2 . 15 . 1 Development Core Team software ( R Foundation for Statistical Computing , Vienna , Austria , 2012 , http://www . R-project . org ) was used to estimate Box-Cox transformation parameter for each cytokine to increase normality of the data . Principal Component Analysis with Varimax rotation was used on all data points simultaneously ( i . e . stimuli AWA/SEA/BCG; total and Treg-depleted PBMC; pre- and post-treatment time-points ) to reduce the data into a smaller number of uncorrelated variables . Rotation converged in 3 iterations and principal components ( PC ) with eigenvalues greater than one were selected . Differences in PC scores between pre- and post-treatment and Treg-depleted and total PBMC were tested with the Wilcoxon matched pairs test . For all tests , statistical significance was considered at the 5% level .
To investigate whether S . haematobium infection affects the frequency of peripheral blood Tregs we compared circulating CD4+CD25hiFOXP3+ Tregs from infected and uninfected children by flow cytometry . Gating strategy for identification of CD4+CD25hiFOXP3+ Tregs is shown in Fig 1A . We found that frequencies of FOXP3+ Tregs were significantly higher in infected children compared to uninfected children ( Fig 1B ) . Importantly , six weeks after praziquantel treatment Treg frequencies were reduced by half to frequencies comparable to the uninfected control group . Over the same six weeks period , there was also a slight but consistent decrease in the Treg frequencies in the uninfected group . Next , we assessed the effect of anthelmintic treatment on cell proliferation and cytokine production in response to stimulation with schistosome-specific antigens SEA and AWA and a non-specific antigen BCG . Proliferation was determined by calculating the division index on the basis of the dilution of CFSE in CD4+CD25hi T cells . There were no significant differences in proliferation between pre-treatment and six weeks post-treatment responses ( medium p = 0 . 397 , AWA p = 0 . 188 , SEA = 0 . 454 and BCG = 0 . 271 ) ( Fig 2 ) . Cytokine production on the other hand significantly changed between pre-treatment and 6 weeks post-treatment; raw cytokine values are shown in S1 Table . We applied Principal Component Analysis ( PCA ) in order to provide a more global assessment of the effect of schistosome infection on responses to not only SEA and AWA stimulation but also to the non-Ag specific stimulant BCG . Two distinct principal components were identified ( Fig 3 and Table 1 ) which captured 73 . 7% of variance in our dataset: principal component 1 ( PC1 ) which reflects regulatory and Th2-polarized cytokine responses due to its positive loading with IL-5 , IL-10 and IL-13 responses ( and accounted for 40% of the total variance in the data ) ; and principal component 2 ( PC2 ) which reflects pro-inflammatory and Th1-polarized cytokine responses due to its positive loading with IFN-γ , IL-17 and TNF ( and accounted for 33 . 7% of total variance in the data . We saw a significant increase in both PC1 and PC2 following treatment compared to baseline values ( Table 2 ) . To study the suppressive effect of Tregs on proliferation and cytokine responses , CD4+CD25hi T cells were depleted from PBMC by magnetic beads . The CD4+CD25hi population decreased by 45% , p = 0 . 0073 of which a representative example is shown in S1B Fig . Depletion of Tregs at pre-treatment resulted in enhanced spontaneous proliferation ( medium condition ) as well as in enhanced proliferation to specific schistosomal antigens AWA and SEA and to vaccine antigen BCG ( Fig 4A ) . At 6 weeks after anthelmintic treatment Treg depletion resulted in significant increase in proliferation in response to AWA only ( Fig 4A ) . A typical plot of CFSE staining showing the effect induced by depletion of Tregs ( Fig 4B ) . Next , we investigated the capacity of Tregs to suppress cytokine responses by evaluating the effect of Treg depletion on principal component 1 ( IL-5 , IL-10 and IL-13 ) and principal component 2 ( IFN-γ , IL-17 and TNF ) . We found that Treg depletion at pre-treatment resulted in increased values of both PC1 and PC2 in the infected individuals , and similarly at post-treatment , the depletion of Tregs resulted in an increase in the values of both PC1 and PC2 in the now infection free schoolchildren ( Table 3 ) .
Down-regulation of immune responses has been attributed to a strong immunomodulatory network of regulatory cells induced by schistosomes [25 , 26] . Here we provide evidence that human Schistosoma infection is associated with increased FOXP3+ regulatory T cells that play a significant role in controlling Th1 and Th2 responses . This finding is consistent with reports of increased numbers of FOXP3+ Tregs in peripheral blood from children 8–13 years old with active schistosomiasis [27] as well as other helminth infections including lymphatic filariasis [14 , 28] . Moreover , the reduction in the number of FOXP3+ Treg after treatment is in line with reports showing that drug induced clearance of Shistosoma parasites reduces Treg numbers defined only as CD4+CD25hi [18] . A much smaller in magnitude , yet statistically significant decrease was observed in the frequency of Tregs in the uninfected control group , which indicates that there were additional factors that affected the measured frequency of the CD4+CD25hiFOXP3+ cells; this could include technical or environmental changes such as seasonal effects , that might be associated with longitudinal studies . However , as both sampling time-points occurred during the long rainy season changes in Treg frequency in the control group are less likely to reflect seasonal changes and other factors such as exposure to concomitant infections could play a role . In order to obtain a global assessment of the effect of S . haematobium infection on Th1 , Th2 , regulatory and pro-inflammatory cytokine responses we applied PCA analysis . This allowed us to summarize the various responses into two principal components [29] . Principal component 1 ( PC1 ) reflected regulatory and Th2-polarized cytokine responses due to its positive loading with IL-5 , IL-10 and IL-13 , responses commonly associated with chronic schistosome infections . Principal component 2 ( PC2 ) reflected pro-inflammatory and Th1-polarized cytokine responses due to its positive loading with IFN-γ , IL-17 and TNF , responses more commonly associated with acute schistosome infection or bacterial infections such as tuberculosis . We show that S . haematobium infection is associated with hypo-responsiveness as demonstrated by increases in cytokine production represented by both PC1 and PC2 following treatment of schistosomiasis with praziquantel . T cell division was also assessed , but despite the consistently higher proliferation to all stimuli tested , at post treatment , the change was not statistically significant . These data indicate that the increased frequency of CD4+CD25hiFOXP3+ Tregs during schistosome infection may be associated with poor cytokine responsiveness . To assess the functional capacity of the regulatory T cells , a field applicable method was used which consists of the depletion of regulatory T cells from PBMC to assess their effect on cytokine production or proliferation . The data show that depletion of Tregs is associated with increased cytokine production , of both PC1 and PC2 which means that both Th2/regulatory and Th1/pro-inflammatory cytokine production improves . This is the case at both pre-treatment and post-treatment time-points , although the increase appears to be stronger at pre-treatment . Altogether , this would suggest that even though regulatory T cell numbers change with infection , their functional capacity to supress cytokine production remains . Furthermore we evaluated the effect of Treg depletion on cell proliferation . While cytokine responses were similarly affected at both pre-and post-treatment , proliferative responses were predominantly affected by Treg depletion in infected individuals at pre-treatment only . These data could be explained if the ability of Tregs to suppress proliferation would be distinct from suppressive mechanisms required for inhibiting cytokine production , for example Tregs only supress PBMC proliferation in a higher Treg:responder ratio as seen during active infection . The removal of S . haematobium infection could then affect the Treg function partially . Tregs are thought to exert their function via a number of different mechanisms including IL-10 and/or TGF-β production , IL-2 consumption , or cell-cell contact where inhibitory molecules such as CTLA-4 and PD-1 are key [30 , 31] . A recent study suggests a unique role for microRNAs in suppression of T cell proliferation by Tregs [32] . Future studies are needed to delineate how Tregs exert their suppressive role during the course of schistosome infection and furthermore the difference in mechanism between suppression of proliferation versus effector cytokine production . Additional alternative mechanism , such as T cell anergy due to increased expression of the E3 ubiquitin ligase GRAIL ( gene related to anergy in lymphocytes ) which has been shown in a mouse model to be linked to Th2 cell hypo-responsiveness could also play a role and should also be investigated in future studies [33] . The role of inflammation in the induction and maintenance of Treg cells should likewise be considered as Shistosoma infection is a chronic inflammatory disorder and Tregs protect the human host against excessive inflammation , thus increased Treg numbers during schistosomisis may be a responses against inflammation rather than directly induced by the parasites [34] . Alternatively , recently described regulatory CD8+ T cells which likewise produce IL-10 , may also in part contribute to the differences observed [35 , 36] . CD25hi cell depletion will in addition to depleting CD4+CD25hiFOXP3+ T cells also deplete the CD8+CD25hiFOXP3+ T cell population and therefore future studies are needed to re-assess the relative contributions of these different subsets . Moreover studies with more extensive panels of markers associated with suppressive T cell functions are necessary as FOXP3 expression has been shown to be transiently up-regulated on activated CD4+ T cells [37] . Finally , the concomitant role of the different regulatory cells , including in addition to Tregs , regulatory B cells [38] and regulatory monocytes/macrophages [39 , 40] and their relative contribution to the suppressive activities observed need to be further investigated . In summary , this study shows that infection with S . haematobium is associated with alterations of the frequency and activity of CD4+CD25hiFOXP3+ regulatory T cells and that these in turn affect proliferation and global cytokine responses . These data indicate that the functional activity of regulatory T cells needs to be taken into consideration when studies consider co-infections , treatment or vaccine responses in areas where helminths are prevalent .
|
Schistosomiasis , a parasitic worm infection , affects over 240 million people worldwide , especially children in sub-Saharan Africa . It is associated with immune hypo-responsiveness which results in an inability of the immune system to eliminate parasites . Animal models suggest that helminths induce regulatory T cells ( Treg ) which suppress effector cells and dampen anti-parasite activity as part of the parasites’ own strategy for survival in the human host . However , little is known about the functional capacity of Tregs during human Schistosoma haematobium infection and their interaction with adaptive responses . We designed a longitudinal study addressing the question of how anti-parasite treatment influences effector T cell activity and Treg function in peripheral blood of schoolchildren living in an S . haematobium endemic area in Lambaréné , Gabon . Our findings show that schistosome infection is associated with increased Treg frequency and that Tregs exert a suppressive effect on immune cell function in terms of both proliferation and cytokine production . Although Treg frequency decreases after anti-schistosome treatment , their suppressive capacity remains unaltered for cytokine production but their influence on proliferation weakens with treatment . By understanding how immune system is prevented from killing parasites , we hope to offer a novel route for intervention to achieve an immunological cure .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
CD4+CD25hiFOXP3+ Regulatory T Cells and Cytokine Responses in Human Schistosomiasis before and after Treatment with Praziquantel
|
Stressful life events are major environmental risk factors for anxiety disorders , although not all individuals exposed to stress develop clinical anxiety . The molecular mechanisms underlying the influence of environmental effects on anxiety are largely unknown . To identify biological pathways mediating stress-related anxiety and resilience to it , we used the chronic social defeat stress ( CSDS ) paradigm in male mice of two inbred strains , C57BL/6NCrl ( B6 ) and DBA/2NCrl ( D2 ) , that differ in their susceptibility to stress . Using a multi-omics approach , we identified differential mRNA , miRNA and protein expression changes in the bed nucleus of the stria terminalis ( BNST ) and blood cells after chronic stress . Integrative gene set enrichment analysis revealed enrichment of mitochondrial-related genes in the BNST and blood of stressed mice . To translate these results to human anxiety , we investigated blood gene expression changes associated with exposure-induced panic attacks . Remarkably , we found reduced expression of mitochondrial-related genes in D2 stress-susceptible mice and in exposure-induced panic attacks in humans , but increased expression of these genes in B6 stress-susceptible mice . Moreover , stress-susceptible vs . stress-resilient B6 mice displayed more mitochondrial cross-sections in the post-synaptic compartment after CSDS . Our findings demonstrate mitochondrial-related alterations in gene expression as an evolutionarily conserved response in stress-related behaviors and validate the use of cross-species approaches in investigating the biological mechanisms underlying anxiety disorders .
Chronic stress is a significant risk factor for human anxiety disorders [1] , yet not all individuals exposed to stress develop a clinically relevant anxiety symptomatology . The underlying reasons for these differences are not yet fully understood but involve an interaction of complex genetic and environmental factors that vary among individuals resulting in stress susceptibility or resilience . The chronic social defeat stress ( CSDS ) model is a well-established mouse paradigm of psychosocial stress , with construct , face , discriminative , and predictive validity for stress-related disorders [2–4] . It involves 10 days of brief daily confrontation of two conspecific male mice , a resident-aggressor and an intruder who reacts with defensive , flight , or submissive behavior [5 , 6] . Although all defeated mice experience stressful stimuli , only some develop stress-related symptoms , measured as social avoidance , making it an excellent model to investigate mechanisms associated with susceptibility and resilience . We have previously shown that genetic factors strongly affect the behavioral outcome of the CSDS , since different inbred mouse strains vary in the proportion of susceptible and resilient animals as well as in their stress coping behaviors [4] . Anxiety disorders are common stress-associated psychiatric disorders [7 , 8] . They are characterized by an excessive physiological and emotional response in the absence of real threat or imminent danger . Among the most debilitating anxiety disorders is panic disorder , which involves sudden recurrent surges of intense fear and discomfort called panic attacks [9] . However , panic attacks are not exclusive to panic disorder , but also frequent in other anxiety disorders . They are typically associated with severe perceived physical and mental stress , feeling of loss of control and avoidance behavior . Epidemiological studies show a major impact of both cumulative and specific life events or stressors , such as threat or psychosocial interpersonal life events , on the development of all anxiety disorders [10 , 11] , including panic disorder [12] . The development of much-needed novel targets for therapeutic intervention of anxiety disorders is limited by the ignorance of the molecular and cellular mechanisms associated with events that initiate and maintain pathological anxiety . The phenotypic heterogeneity of human populations and the high variability of environmental influences [13] , along with a limited access to brain tissue samples , make it difficult to identify the biological basis of anxiety disorders . These challenges can , to some extent , be controlled in animal models . Cross-species approaches are therefore expected to reveal specific biological mechanisms underlying anxiety disorders [14] . They can be especially powerful for multi-omics studies allowing hypothesis-free identification of the most significant biological pathways associated with specific exposures , while at the same time probing the effects of different genetic backgrounds on the outcomes . Therefore , the integration of multiple levels of information [15] and the translation of the results from animal models to the human condition are critical for the success of cross-species approaches . Omics-based approaches require the investigation of etiologically relevant tissue . The CSDS model has been used to study transcriptomic effects of chronic psychosocial stress in brain regions classically associated with anxiety and depression-related behaviors , including the medial prefrontal cortex , hippocampus , amygdala , and nucleus accumbens [4 , 16] . In addition , converging lines of anatomical , physiological , and mechanistic evidence suggest that the bed nucleus of the stria terminalis ( BNST ) functions as the center of integration for limbic information [17] , monitoring the genesis of long-term responses to anxiety [18] , such as anticipatory anxiety . The BNST is an especially relevant brain region for stress-related anxiety . It has been suggested that it plays a role in valence surveillance by processing salient information on physical and social contexts , collected through its numerous projections throughout the brain [17] . However , only few anxiety studies have generated large-scale genomics data from the BNST [19] . To identify the core differentially expressed molecules and pathways underlying pathological anxiety and resilience to it , we employed a multi-omics approach in mice . We applied the CSDS model to induce anxiety-related phenotypes and to identify molecular markers for susceptibility and resilience in male mice . We used two inbred strains , the largely stress-resilient C57BL/6NCrl ( B6 ) and stress-susceptible DBA/2NCrl ( D2 ) , due to strong genetic background effects in the mouse stress response [4 , 20] . We investigated gene ( RNA-seq ) , microRNA ( miRNA-seq ) , and protein ( LC-MS/MS ) expression differences in the BNST between the stress-resilient , stress-susceptible , and control mice . As a translational effort , we also carried out RNA-seq from mouse post-CSDS blood , and longitudinal microarray-based gene expression profiling from blood cells of panic disorder patients who experienced high degrees of stress and anxiety due to exposure to phobic situations . To identify converging anxiety-related gene sets and pathways , we conducted pathway and gene set enrichment analyses of the data sets from mouse BNST samples , and mouse and human blood cells . Altogether , our results indicate anxiety induces a genetically-controlled evolutionarily conserved response in mitochondrial pathways .
To study how genetic background affects the behavioral response to chronic psychosocial stress , we subjected B6 and D2 mice to a 10-day CSDS , followed by the social avoidance ( SA ) test 24 hours later ( Fig 1 , S1 Table ) . Since these strains differ in their innate social avoidance levels [4] , we evaluated the behavior of the defeated mice in comparison to the same-strain controls . We divided the defeated group of animals and defined stress-susceptible mice as those with social interaction ( SI ) ratios below one standard deviation from the mean SI ratio of the same-strain controls , as previously described [4] . We classified all other defeated mice with ratios above those values , i . e . resembling controls , as resilient ( Fig 1E ) . The strains showed distinct response to stress since 89% of D2 mice , but only 30% of B6 mice , presented social avoidance behavior , being susceptible to chronic psychosocial stress ( Pearson’s chi-square , χ2 = 60 . 38 , P = 7 . 76E-14 ) ( Fig 1F ) . CSDS also significantly affected locomotor behavior of mice . Both B6 and D2 susceptible mice moved significantly less than same-strain controls ( P = 0 . 003 and P = 0 . 002 , respectively ) during the no-target trial , i . e . , the trial without the social target mouse ( Fig 1D ) , of the SA test . D2 resilient mice moved significantly more than D2 susceptible mice ( P = 0 . 003 ) , while no such difference was observed in the B6 strain ( P = 0 . 443 ) ( S1 Table ) . To assess if chronic psychosocial stress affects the duration to cease escape-oriented behavior in the face of an acute stressor , we performed the forced swim test ( FST ) five days after the end of CSDS . The latency to immobility during the FST , used as a measure of active stress coping [21 , 22] , was highly correlated with the SI ratio in the D2 defeated mice ( r = 0 . 920 , P = 0 . 009 ) ( S1 Table ) . In other words , defeated D2 mice with higher resilience to psychosocial stress presented a more active coping strategy than mice with higher social avoidance . We did not observe similar correlations in the D2 control group ( r = -0 . 130 , P = 0 . 759 ) or in the B6 control or defeated mice ( r = -0 . 079 , P = 0 . 691 and r = 0 . 026 , P = 0 . 852 , respectively ) ( S1 Table ) . We also found that CSDS had a significant effect on the body weight of the D2 , but not the B6 mice ( S1 Fig ) . Overall , these results suggest that genetic background has a strong effect on stress susceptibility , with the D2 strain being more susceptible to stress-induced social avoidance than the B6 strain . Furthermore , the two strains used different strategies to cope with stress , as demonstrated by their differences in locomotor and escape-oriented behavior after , and body weight development throughout , the CSDS . To identify stress-associated transcriptomic and proteomic signatures in the B6 and D2 mouse strains , we profiled the BNST , a key brain region regulating anxiety . Profiling was conducted approximately one week following completion of the CSDS , when we sacrificed all mice and dissected BNSTs for analyses ( Fig 2A and 2B ) . We then carried out both gene expression ( RNA-seq ) and proteomic ( liquid chromatography-tandem mass spectrometry ) profiling to identify differentially expressed mRNAs ( data set A ) and proteins ( data set B ) ( Fig 2C ) . Additionally , in the B6 strain , we performed Argonaute 2 RNA immunoprecipitation-sequencing ( AGO2 RIP-seq ) of active microRNAs ( miRNAs ) and their mRNA targets ( data sets D and E , respectively ) . AGO2 is the catalytic component of the RNA-induced silencing complex ( RISC ) . Only mature miRNAs can be incorporated into RISC in the presence of AGO2 , and serve as a guide molecule for silencing their target mRNAs [23] . Thus , AGO2 RIP-seq identifies only those miRNAs and their target mRNAs that are bound to the RISC at the time of tissue collection , providing additional specificity compared to the sequencing of all cellular miRNAs and mRNAs . Data sets A and B were collected from the same cohorts of animals , with each cohort being equally divided by the SI ratios between transcriptomic and proteomic experiments . AGO2 RIP-seq ( data set D and E ) was performed from an additional cohort . In all data sets , we compared the stress-resilient , stress-susceptible , and same-strain control mice . For the resilient and susceptible groups , we selected mice representing the phenotypic extremes , i . e . , those with the highest and lowest SI ratios , respectively . Unless stated otherwise , “differentially expressed” ( DE ) mRNAs , miRNAs and proteins were defined as P < 0 . 05 and |FC| ≥ 1 . 2 . For all individual findings of DE mRNAs , miRNAs and proteins , we report both nominal P-values and P-values adjusted for multiple testing by the modified Benjamini–Hochberg method ( Padj ) as defined in [24] . The total numbers of genes and proteins DE at P < 0 . 05 and Padj < 0 . 05 levels for each data set are presented in S2 Table . We next asked if the same pathways and gene sets that we found to be stress-responsive in the mouse BNST , could be identified from an accessible tissue , i . e . blood cells , and whether these pathways and gene sets were also DE in blood of humans with anxiety . One week after CSDS , we collected blood samples from stress-susceptible , resilient , and control B6 and D2 mice , and carried out RNA-seq and miRNA-seq ( data sets F and G , respectively ) . Blood samples were collected from the same mice that were used for BNST RNA-seq ( data set A ) , except for D2 resilient mice where we did not obtain a sufficient amount of blood . As a translational human anxiety disorder data set , we collected samples from panic disorder patients who underwent an exposure intervention . We selected panic disorder as our translational target to concentrate on a phenotypically homogeneous sample . We collected blood samples at baseline , 1 h after anxiety peak during exposure and 24 h after exposure-induced panic attack , and carried out microarray-based gene expression profiling . To determine if the observed differences in mitochondrial gene and protein expression are associated with changes in mitochondrial morphology in the pre-synaptic or post-synaptic compartments of neurons , we carried out transmission electron microscopy ( TEM ) in the BNST after CSDS in B6 and D2 mice . We classified mitochondria as “synaptic” if the synaptic density and vesicles within the post-synaptic terminal were clearly visible ( Fig 8A ) . We observed stress-associated differences in maximum ( length ) and minimum ( width ) diameters of mitochondrial cross-sections in the B6 but not in the D2 strain . The B6 susceptible mice had on average 8 . 4% shorter mitochondrial cross-sections ( maximum diameter ) than B6 controls ( Padj = 0 . 003 ) ( Fig 8B ) , but no differences in the width ( Fig 8C ) . However , the mean mitochondrial cross section length/width ratio in D2 susceptible , but not B6 , mice was 5% larger than in the resilient group ( Padj = 0 . 003 ) indicative of increased maximum , but decreased minimum , diameter ( Fig 8D ) . The mean number of mitochondrial cross-sections was not influenced by stress ( Fig 8E ) . However , when we investigated the pre-synaptic and post-synaptic compartments separately ( Fig 8F and 8G ) , we detected 39% more pre-synaptic cross-sections in susceptible compared to the control B6 mice and 46% less post-synaptic cross-sections in resilient compared to the susceptible B6 mice . In addition , we observed significant strain differences . In general , B6 mice had wider diameter and smaller number of mitochondrial cross-sections than D2 mice , both pre- and post-synaptically ( Fig 8B , 8F and 8G ) . Thus , consistently with our gene and protein differential expression data , we observed significant strain-dependent changes in mitochondrial morphology in the BNST following CSDS .
To identify the key biological pathways mediating resilience and susceptibility to psychosocial stress , a risk factor for onset and recurrence of anxiety disorders , we applied a cross-species multi-omics approach . In a chronic psychosocial stress mouse model , we found differential expression of mitochondrial-related genes and proteins both in the BNST and blood cells . However , the pattern of differential expression was opposite in the B6 and D2 mouse strains . Subsequently , we tested whether the same pathways are involved in acute anxiety provocation in panic disorder patients . Our analyses revealed a consistent convergence of differentially expressed mitochondria-related pathways in the blood samples from panic disorder patients after exposure-induced panic attack . As in the stress-susceptible D2 mice , these genes were downregulated during and after panic attack in patients . Consistently , we observed significant strain-dependent stress-associated differences in mitochondrial morphology in the BNST . Taken together , our results have uncovered an evolutionarily-conserved mitochondrial signature that characterizes anxiety-related behavior in mammals . Of the mitochondrial genes , especially those related to oxidative phosphorylation were differentially expressed in both BNST and blood cells after chronic stress in mice and during exposure-induced panic attacks in panic disorder patients . These genes , that regulate both ATP production and apoptosis , had lower expression levels in the susceptible D2 mice compared to controls and in the panic disorder patients during and after panic attack . The expression levels of the same genes were higher in the susceptible B6 mice compared to controls . In a bidirectionally bred mouse model of trait anxiety ( the HAB/LAB mice ) , we previously observed increased expression of electron transport chain proteins in the cingulate cortex synaptosomes of the high-anxiety mice [47] . In an outbred strain rat model of social behavior , highly anxious rats that were prone to become subordinate during a social encounter with a rat with low levels of anxiety had lower levels of mitochondrial complex I and II proteins in the nucleus accumbens [48] . In a study that specifically investigated gene expression of mtDNA-encoded genes [49] , four of these genes ( mt-Nd1 , mt-Nd3 , mt-Nd6 , and mt-Atp6 ) were downregulated after acute immobilization stress in the hippocampus . However , after chronic immobilization stress mt-Nd6 was upregulated . These effects were mediated by glucocorticoids . Thus , also previous studies have found changes in brain mitochondrial gene expression in anxiety-like behavior and after stress , but the directionality of these changes was depended on the model of anxiety and the duration of the applied stressor ( e . g . acute versus chronic stress ) . Our results extend these previous observations by showing that the directionality of the changes may likely be influenced by the genetic background of the strain , and may be related to the innate anxiety level or stress susceptibility of these strains . Moreover , the greatest advantage of multi-omics studies is their ability to identify the most significantly affected pathways from measurements of thousands of molecules . Although mitochondrial pathways have previously been associated with anxiety , our hypothesis-free approach established their dysregulation as a major brain stress response . We also observed differences in mitochondrial morphology and/or number of mitochondrial cross-sections in B6 and D2 mice after CSDS in the BNST . Stress-susceptible B6 mice had a larger number but shorter pre- and post-synaptic mitochondrial cross-sections compared to B6 control or resilient mice . In the D2 strain , stress-susceptible mice had slightly increased mitochondrial cross-section length/width ratio compared to the resilient mice . Previously , nocturnal aggression stress has been associated with slightly smaller pineal gland mitochondrial size in gerbils [50] . Chronic , but not acute , immobilization stress in rats leads to increased mitochondrial area in hippocampal mossy fiber terminals [51] . Thus , psychological stress is associated with morphological changes of mitochondria , but how these changes relate to mitochondrial function remains to be revealed . Cellular stress can affect mitochondrial morphology , e . g . in the form of hyperelongation , donut formation , or fragmentation in response to cytochrome c release [52] . In response to ER stress , which leads to downregulation of protein synthesis through the eIF2 pathway involved in mRNA translation , mitochondria reshape and become longer to promote cellular energy levels [53] . It has been proposed that the cumulative effect of stressors over time contribute to mitochondrial allostatic load and overload , and as a consequence lead to recalibration of mitochondrial structure and functional adaptation ( e . g . activation of hormonal receptors ) as well as dysregulation of gene expression , inflammatory response , and apoptosis [54–56] . In this model , mitochondria interact bidirectionally with stress mediators , but their adaptive capacity and the direction of changes related to mitochondrial function in response to stressors depend on the innate resilience of the organism and the effects of long-term programming during critical developmental periods . In addition to the opposite directionality of mitochondria-related pathways , we found strain-specific expression patterns of differentially expressed miRNAs . Notably , miR-34c was expressed at a lower level in both blood cells and BNST of B6 resilient mice compared to controls after CSDS . Several members of the miRNA-34 family are differentially expressed in psychiatric conditions in humans: miR-34a is expressed at a higher levels in blood cells of patients with schizophrenia [57] , and miR-34b-5p and mir-34c-5p in patients of major depressive disorder [58] . Furthermore , miR-34a expression is higher in the postmortem cerebellum of patients with bipolar disorder [59] . In mice , we have previously shown that both miR-34a and miR-34c are induced by acute and chronic stress in the amygdala , and that injection of miR-34c to amygdala is anxiolytic after a stress challenge [60] . In a rat model of early adolescent traumatic stress , miR-34c was upregulated in the hypothalamus [61] . Many of the miR-34 family effects on stress may be mediated through the CRFR1 [60] . We did not find CRFR1 differential expression in the BNST or blood , suggesting existence of alternative main targets in these tissues . In addition to these two miRNAs differentially expressed only in the B6 strain , we found two out of three miRNAs , miR-148a and miR-181b , to be DE in opposite directions between the strains in blood cells . miR-148a has been previously associated with panic disorder [45] and miR-181b has been identified as a possible marker for schizophrenia [46] and Alzheimer’s disease [62] . Higher expression of miR-181b has been previously correlated with an increase in mitochondrial oxidative stress and oxidative DNA damage [63] , an early and systemic process in the pathophysiology of Alzheimer’s disease [64 , 65] . Both miRNAs are highly conserved between mice and humans [66] . In addition to differential expression of mitochondria-related pathways and miRNAs , we found distinct molecular responses in the two strains in dopamine- and cAMP-regulated neuronal phosphoprotein ( DARPP-32/PPP1R1B ) and associated Dopamine-DARPP-32 Feedback in cAMP and Ca2+ Signaling pathway . Increased expression of DARPP-32 protein was previously observed in the prefrontal cortex and the amygdala of socially defeated B6 mice [36] . DARPP-32 has been proposed to act as an integrator of dopaminergic and glutamatergic signaling [67] and elevated levels of its truncated isoform were reported in schizophrenia , bipolar disorders and major depression as well as poor cognitive functioning [40] . These results implicate genetic background-dependent differences in susceptibility to chronic stress between the strains [68 , 69] . Overall , our results suggest that transcriptomic response to stress is strongly dependent on the genetic background . At the behavioral level , innately anxious D2 mice were more susceptible to CSDS than the less anxious B6 mice . While relatively few studies have examined strain differences in response to repeated stress [68 , 70–72] , and CSDS in particular [4 , 73 , 74] , their findings similarly implicate heightened stress susceptibility in more anxious strains ( e . g . , BALB/cJ , D2 ) . D2 resilient mice responded to CSDS with higher stress-induced locomotor activity in comparison to the D2 susceptible mice . Additionally , D2 resilient mice with higher social interaction ratios also had an elevated latency to immobility in the FST , suggesting resistance to behavioral despair or lack of adaptation aimed at energy conservation in the face of inescapable situation [75 , 76] . In contrast , B6 mice did not differ in either measurement . The two strains also varied in the body weight development during and after CSDS . While the body weight decreased in the defeated D2 mice in comparison to the D2 controls and the baseline weight , the defeated B6 mice gained weight during the CSDS , similarly to the B6 controls . Taken together , we found that genetic background influences adaptation of different stress-coping strategies , as observed previously [4 , 77] . For neuropsychiatric diseases , such as panic disorder , obtaining samples from the brain is essentially impossible for a reasonably large and representative sample sets . Therefore , the existence of tissue , which is more accessible and could be used as surrogate for gene expression in the central nervous system , is crucial . In our data set , the overall gene expression pattern between the BNST and blood cells was moderately correlated in both mouse strains . These results are in accordance with previous studies investigating gene expression similarities between the whole blood and multiple brain regions in humans , and suggest that gene expression is not perfectly correlated between brain and blood , but may be useful for studying certain pathways , e . g . related to translational control [78] . On the pathway level , oxidative phosphorylation-related genes were differentially expressed both in the mouse BNST and blood cells , and in the panic disorder patient blood cells . These genes were upregulated in the defeated B6 and downregulated in the defeated D2 mice compared to controls both in the BNST and blood cells . Strikingly , this pathway was downregulated in panic disorder patients directly and 24 h after an exposure-induced panic attack . It is interesting that the gene expression pattern of the patients resembles that of the more stress-susceptible mouse strain , suggesting that stress-susceptibility may involve a general modulation of genes associated with mitochondrial function . There are several caveats in comparing a post-mitotic brain region to a peripheral biospecimen , which are very different in terms of mitochondrial metabolism , but nonetheless , we observed a consistent and converging signature that warrants further investigations . Altogether , our data reinforce the utility of cross-species approaches in the identification of the biological basis of human anxiety disorders but advises careful selection of mouse strains for the best translatability of the findings . Although we concentrated on mitochondrial pathways and anxiety susceptibility in our follow-up experiments , we also found several pathways in the mouse transcriptomic and proteomic experiments associated with stress resilience . Stress resilience is considered an active , adaptive response to stress , not merely a lack of maladaptive symptoms [79] . Studying of resilience is challenging in humans due to a lack of well-controlled cohorts . However , translational studies on stress resilience would be highly warranted to understand the underlying mechanisms providing putative means to enhance resilience in stress-susceptible individuals . In conclusion , our cross-species multi–omics approach found a systemic evolutionarily conserved mitochondrial response in anxiety-related behaviors in mice and in panic disorder patients . We have produced a large amount of mRNA , miRNA , and protein expression data and made it publicly available to provide a resource for formulation of additional hypotheses on psychosocial stress-induced anxiety in mice , and panic disorder in humans . Our converging findings on stress-susceptibility in both brain and blood cells indicated dysregulation of translation and mitochondrial-related pathways . Further functional studies underlying the mechanisms behind the observed differences in mitochondrial morphology and the dysregulation of the mitochondria-related genes will provide much needed insight into the molecular basis of panic disorder and other anxiety disorders , a critical step in developing future targeted therapies .
All animal procedures were approved by the Regional State Administration Agency for Southern Finland ( ESAVI-3801-041003-2011 and ESAVI/2766/04 . 10 . 07/2014 ) and carried out in accordance with directive 2010/63/EU of the European Parliament and of the Council , and the Finnish Act on the Protection of Animals Used for Science or Educational Purposes ( 497/2013 ) . All human procedures and research were approved by the Ethics Committee of the Ludwig Maximilian University , Munich , Germany , Project MARS Anxiety , IRB Number 318/00 , in accordance with the Declaration of Helsinki . Written informed consent was obtained from all participants . Male 5-week-old B6 and D2 , and 13-26-week-old Clr-CD1 ( CD-1 ) mice ( Charles River Laboratories , Sulzfeld , Germany ) were housed in pathogen-free , humidity- ( 50 ± 15% ) and temperature-controlled ( 22 ± 2°C ) animal facility on a 12 h light-dark cycle ( lights on at 6 a . m . ) and with ad libitum access to Teklad 2916 rodent chow ( Envigo , Huntingdon , United Kingdom ) and water . All B6 and D2 mice were acclimatized for 10 days prior to CSDS in groups of 4–6 mice housed in Makrolon Type III cages . CD-1 mice were acclimatized for one week in single individually ventilated cages ( IVC ) ( Tecniplast , Buguggiate , Italy ) prior to CD-1 aggressor screening . Twenty-one ( n = 6 males , age 29 . 33 ± 8 . 48 years; n = 15 females , age 32 . 60 ± 9 . 61 years ) non-medicated panic disorder patients were recruited in the anxiety disorder outpatient unit at the Max Planck Institute of Psychiatry , Munich , Germany . Panic disorder without ( n = 3; 14 . 3% ) or with ( n = 18; 85 . 7% ) comorbid agoraphobia was assigned as the primary diagnosis , mild secondary depression was allowed ( n = 2; 9 . 5% ) . The diagnosis was ascertained by trained psychiatrists according to the Diagnostic and Statistical Manual of Mental Disorders ( DSM ) -IV criteria as previously described [96] . Exposure sessions were conducted outside the clinic , depending on the feared situation ( e . g . subway , supermarket , tower ) and specific concern ( e . g . fainting , asphyxiation , losing control ) . The kind and site of exposure was determined by the patient , with the goal of highest fear provocation . On the day of the exposure and post-exposure day , the patients were instructed to eat a certain breakfast; smoking , exercises/sports or intake of caffeine was not allowed . Each exposure was performed in the morning , starting at 8 a . m . to 9 a . m . For all three time points ( baseline , 1 h and 24 h post-exposure ) peripheral blood was collected using PAXgene Blood RNA Tubes ( PreAnalytiX , Hombrechtikon , Switzerland ) and processed as previously described [96] . Blood cell RNA was hybridized to Illumina HumanHT-12 v4 Expression BeadChips ( Illumina , CA , USA ) . Raw probe intensities were exported with Illumina’s GenomeStudio . Cross-hybridizing probes as well as probes binding to X and Y chromosomes were removed to avoid a possible gender effect [96] . Probes with detection P-value larger than 0 . 05 in > 50% of the samples were excluded from the analysis . For each transcript , normalization was performed using Variance stabilization and calibration for microarray data ( VSN ) R package . Subsequently , technical batches associated to Chip Barcode and Bead Chip ID were identified and removed with ComBat [91] . The probes were annotated to HGNC GeneSymbols using the Illumina platform annotation file and biomaRt R package v2 . 36 . 1 [97] . Probes not annotated to any genes and those annotated to multiple genes were excluded from downstream gene set enrichment analyses . Gene expression data is available in GEO ( GSE119995 ) . To select proteins for technical validation with Western blot , we performed a literature search on molecules overlapping the transcriptomic ( A and E ) and proteomic ( B ) data sets , and which were also present in at least one of the significantly dysregulated canonical pathways ( Fig 4A ) or mitochondria-related gene sets ( S3 Fig ) , and found nine proteins ( ADCY5 , PPP1R1B , QDPR , GAD2 , ATP2B1 , PPP3CA , ATP6V1E1 , GLUD1 , and CYCS; Fig 5 ) [100] . We selected two , PPP1R1B and CYCS , which have been previously associated with psychiatric disorders [36–39] , for validation . Validation samples included a subset of four or five samples used in LC-MS/MS analysis ( data set B ) , with the exception of all D2 resilent mice being identical . Ten μg of total protein were separated by SDS-PAGE and transferred to Immobilon-FL membranes ( Merck Millipore , MA , USA ) . After blocking with 5% non-fat milk in Tris-buffered saline with 0 . 1% Tween-20 ( TBS/T ) , membranes were incubated overnight at 4°C with a primary antibody: mouse monoclonal anti-Cytochrome C antibody ( 1∶500 , Santa Cruz Biotechnology , TX , USA , #sc13156 ) or mouse monoclonal anti-DARPP-32 antibody ( 1∶200 , Santa Cruz Biotechnology #sc-271111 ) . After washing , membranes were incubated with horseradish peroxidase ( HRP ) -conjugated secondary antibody ( goat anti-rabbit , 1∶10000 , Cell Signaling Technology , MA , USA , #7074S or goat anti-mouse , 1∶10000 , Santa Cruz Biotechnology , #sc-516102 ) . Signals were visualized with the Chemi Doc MP Imaging System ( BioRad , Munich , Germany ) after incubating membranes with enhanced chemiluminescence developing solution ( Merck Millipore , Darmstadt , Germany ) . Expression levels of all proteins were normalized to Coomassie blue staining signals . Densitometric analyses were carried out using ImageJ software ( National Institutes of Health , MD , USA ) . Statistical significance was calculated between data pairs with a one-tailed Student t test using Microsoft Excel .
|
Genetic and environmental factors contribute to the etiology of psychiatric diseases but the underlying mechanisms are poorly understood . Chronic psychosocial stress is a well-known risk factor for anxiety disorders . To identify biological pathways involved in psychosocial stress-induced anxiety and resilience to it , we used a well-characterized mouse model of chronic social defeat stress ( CSDS ) in two inbred mouse strains , C57BL/6NCrl ( B6 ) and DBA/2NCrl ( D2 ) , which differ in their susceptibility to stress . We focused on the bed nucleus of the stria terminalis , a key brain region behind stress-response and anxiety , and carried out genome-wide analysis of mRNA , and miRNA expression , and protein abundance . Bioinformatic integration of these data supported differences in mitochondrial pathways as a major stress response . To translate these findings to human anxiety , we investigated blood cell gene expression in mice and in panic disorder patients exposed to fearful situations and experiencing panic attacks . Concurring with our brain findings , expression of mitochondrial pathways was also affected in mouse and human blood cells , suggesting that the observed stress response mechanisms are evolutionarily conserved . Therefore , chronic stress may critically affect cellular energy metabolism , a finding that may offer new targets for therapeutic interventions of stress-related diseases .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"blood",
"cells",
"neuropsychiatric",
"disorders",
"medicine",
"and",
"health",
"sciences",
"anxiety",
"disorders",
"natural",
"antisense",
"transcripts",
"gene",
"regulation",
"vertebrates",
"social",
"sciences",
"mice",
"animals",
"mammals",
"micrornas",
"genome",
"analysis",
"mitochondria",
"bioenergetics",
"cellular",
"structures",
"and",
"organelles",
"genomics",
"neuroses",
"animal",
"cells",
"gene",
"expression",
"mental",
"health",
"and",
"psychiatry",
"biochemistry",
"rna",
"psychology",
"rodents",
"eukaryota",
"cell",
"biology",
"panic",
"disorder",
"nucleic",
"acids",
"genetics",
"transcriptome",
"analysis",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"energy-producing",
"organelles",
"non-coding",
"rna",
"amniotes",
"computational",
"biology",
"organisms",
"psychological",
"stress"
] |
2019
|
Multi-omics analysis identifies mitochondrial pathways associated with anxiety-related behavior
|
We computationally determined miRs that are significantly connected to molecular pathways by utilizing gene expression profiles in different cancer types such as glioblastomas , ovarian and breast cancers . Specifically , we assumed that the knowledge of physical interactions between miRs and genes indicated subsets of important miRs ( IM ) that significantly contributed to the regression of pathway-specific enrichment scores . Despite the different nature of the considered cancer types , we found strongly overlapping sets of IMs . Furthermore , IMs that were important for many pathways were enriched with literature-curated cancer and differentially expressed miRs . Such sets of IMs also coincided well with clusters of miRs that were experimentally indicated in numerous other cancer types . In particular , we focused on an overlapping set of 99 overall important miRs ( OIM ) that were found in glioblastomas , ovarian and breast cancers simultaneously . Notably , we observed that interactions between OIMs and leading edge genes of differentially expressed pathways were characterized by considerable changes in their expression correlations . Such gains/losses of miR and gene expression correlation indicated miR/gene pairs that may play a causal role in the underlying cancers .
MicroRNAs ( miRs ) are small non-coding RNAs that interact with their gene target coding mRNAs . Such small RNAs putatively inhibit translation by direct and imperfect binding to the 3′- and 5′-untranslated regions ( UTR ) [1] and exert expression control with other regulatory elements such as transcription factors [2] , [3] , [4] . The elementary role of miRs in gene expression has been indicated in tissue- and organ-specific development [5] . miRs also play an important role in tumors [6] , [7] , [8] , where over-expressed miRs might diminish the level of expression of targeted tumor suppressor genes [9] . In turn , miRs may act as tumor suppressors , when their down-regulation leads to enhanced expression of targeted oncogenes [10] or are involved in various steps of the metastatic process [11] . Generally , aberrant expression of miRs in cancers can arise from the deletion or mutation as well as methylation of miR coding regions [12] . Furthermore , miRs may be located in common breakpoint regions and genomic areas of amplification and loss of heterozygosity [13] . Such alterations of miR-expression levels have been implicated in the de-regulation of critical players in major cellular pathways , modifying the differentiation , proliferation and survival of tumor cells . For example , miR-7 and miR-221/222 have been shown to be involved in the activation of the Akt and epidermal growth factor receptor ( EGFR ) signaling pathways in gliomas [14] , [15] while miR-34a was found to be a key regulator of p53 [16] . To provide a better understanding of the involvement of miRs in pathways , we computationally determined miRs that are significantly associated with molecular pathways . In particular , we utilized gene expression profiles to determine a pathway specific enrichment score in diverse cancer types , such as glioblastomas , ovarian and breast cancers . Using data of physical interactions between miRs and the 3′UTR of mRNAs we counted the numbers of leading edge genes ( LEG ) in each pathway that were targeted by a given miR . We assumed that the topology of interactions between LEGs of pathways and miRs allows an assessment of the tumor-specific importance of the given miR for the expression of the underlying pathways . Therefore , we used a machine learning approach to fit pathway-specific enrichment scores as a function of the corresponding number of LEGs that were targeted by an array of miRs . Despite the diversity of the underlying cancer types , we obtained a large , overlapping set of important miRs ( IM ) that significantly influenced the regression process in all cancer types considered . Furthermore , IMs that were important for an increasing number of pathways were enriched with literature curated cancer miRs and differentially expressed miRs . Such sets of IMs also coincided well with clusters of miRs that were experimentally indicated in numerous other cancer types . Focusing on such an overlapping set of overall important miRs ( OIM ) in glioblastomas , ovarian and breast cancers , we investigated their interactions to LEGs in differentially expressed pathways . We observed that such interactions were characterized by considerable changes in their expression correlations . Such gains or losses of expression correlations indicated OIM/LEG pairs that may influence expression changes in the underlying pathways .
Using The Cancer Genome Atlas ( TCGA , http://cancergenome . nih . gov/ ) , we utilized 77 glioblastoma samples and 10 non-tumor control samples that provided matching gene and miR expression profiles . We also used 77 samples of ovarian cancer and 8 non-cancer tumor samples , as well as 79 breast cancer and 19 non-cancer control samples . Comparing disease and control samples , we determined differentially expressed miRs by a Student's t-test if FDR<0 . 01 . Accordingly , we found 164 differentially expressed miRs in GBMs , 282 in ovarian and 82 in breast cancers . We collected overlapping sets of 35 oncomiRs , 42 tumor suppressor- miRs [6] , [9] , [17] , [18] , [19] , [20] , [21] and 32 miRs that were involved in metastasis [11] , [19] , [20] , [21] , [22] ( Fig . S1 ) . The HMDD database [23] collects reports from the literature that experimentally indicated a miRs involvement in different tumor types . Specifically , we utilized sets of 45 miRs in glioblastomas , 81 in ovarian and 125 in breast cancer . As a source of reliable protein pathway information , we used 429 annotated pathways from the Reactome database [24] . Utilizing human specific data from PicTar [25] , miRanda [26] , [27] and TargetScanS [28] we assembled 48 , 939 interactions between 386 miRNAs and 6 , 725 mRNAs , demanding that each interaction was reported by at least two sources [29] . All interaction pairs are presented in Table S1 . Using gene expression data of a cancer type , we applied GSEA [30] to calculate a normalized enrichment score of each pathway . We represented each pathway by a profile of miRs that reflected the number of leading edge genes ( LEG ) in the underlying pathway a given miR interacts with . Focusing on a given miR we normalized such numbers by a Z-score averaging over all pathways . Finally , we used random forest algorithm [31] to perform a regression of the pathways normalized enrichment scores as a function of the miR profiles of Z-scores . In each of 10 , 000 regression trees , we randomly sampled of all n miRs and of all x pathways [21] , [29] . As for the assessment of a miR's importance for each pathway in the fitting process , we permuted enrichment scores and the number of targeted LEGs , calculating randomized local importance values for each miR/pathway pair . We repeated the randomization process 100 times and constructed null-distributions of randomized importance scores for each miR/pathway pair . Fitting such distributions with a Z-test , we calculated P-values for each miR/pathway pair . We corrected for multiple testing by calculating the corresponding false discovery rate ( FDR ) [32] and defined an important miR ( IM ) of a pathway if FDR<0 . 01 . We grouped important miRs ( IM ) according to their number of pathways . Specifically , we represented each group by IMs that had at least k pathways . In each group we calculated the number of IMs with a certain feature i ( i . e . being differentially expressed or a cancer miR ) , . Randomly assigning feature i to IMs we defined as the enrichment of IMs with feature i where was the corresponding random number of IMs with feature i among all IMs . After averaging Ei over 10 , 000 randomizations Ei>1 pointed to an enrichment and vice versa , while Ei∼1 indicated a random process [33] . Analogously , we determined the enrichment of differentially expressed pathways as a function of the number of their IMs . Assuming ND cancer and NC non-tumor control samples , we calculated Pearson's correlation coefficient of an interacting miR i and gene j in the disease ( ) and control ( ) samples . Subsequently , we Fisher transformed correlation coefficients into a Z-score reflecting the difference of correlation coefficients defined as . Therefore , a positive ΔZ corresponded to a gain of correlation in the disease case and vice versa .
In the first step of our procedure ( Fig . 1A ) , we applied Gene Set Enrichment Analysis ( GSEA ) [30] to determine a normalized enrichment score of each pathway , comparing expression profiles in cancer cases to their non-cancer controls . Accounting for the expression characteristics of different cancer types , we represented each pathway by ‘leading edge genes’ ( LEG ) , a subset of genes that significantly drove the enrichment of a given pathway in the disease cases [30] . Furthermore , we assembled 48 , 939 interactions between 386 miRs and 6 , 725 mRNAs ( Table S1 ) . Pooling such miR-gene interaction data from PicTar [25] , miRanda [26] , [27] and TargetScan [28] , we demanded that each interaction was reported by at least two sources [29] . Considering each pathway as a set of LEGs , we counted the number of such genes that a given miR interacted with . Consequently , each pathway was further represented by a miR interaction profile , indicating the number of LEGs in a pathway a given miR interacted with ( Fig . 1B ) . Averaging over all pathways , we normalized miR-specific entries in this matrix by a Z-score . Representing each pathway by its normalized enrichment score , we applied the random forest algorithm , allowing the calculation of an importance value for each miR/pathway pair . Such an importance measure reflects the impact of the given miR on the fitting process of the underlying pathway's enrichment score . To assess the statistical significance of local importance scores we resorted to permutation tests ( Fig . 1C ) . Randomizing both pathway enrichment scores and the miRs numbers of targeted LEGs we generated null-distributions of importance scores for each miR/pathway pair . Utilizing a Z-test we determined P-values and observed a pair of an important miR ( IM ) and a pathway if FDR<0 . 01 [32] . In glioblastomas , we found a total of 2 , 320 significant pairs between 167 IMs ( 49 . 6% out of all miRs that interacted with LEGs in 429 pathways ) and 265 pathways ( 61 . 8% out of all 429 pathways ) . Furthermore , we observed that the set of pathways was significantly enriched with differentially expressed pathways as provided by GSEA ( FDR<0 . 01 ) applying Fisher's exact test ( P<10−12 ) . Similarly , we found 2 , 564 pairs between 171 IMs ( 50 . 3% ) and 322 pathways ( 75 . 1% ) in ovarian cancer ( P<10−7 ) while 156 IMs ( 47 . 3% ) were linked to 309 pathways ( 72 . 0% ) through 2 , 041 pairs in breast cancer ( P<10−9 ) . For a complete list of all IM/pathway pairs see Tables S2 , S3 , S4 . In Fig . 1D , we observed that sets of IMs largely overlapped , allowing us to find 99 overall important miRs ( OIM ) , corresponding to 59 . 2% of IMs in GBM , 57 . 9% in ovarian and 63 . 5% in breast cancer . In turn , we also found that pathways overlapped strongly ( Fig . S2A ) with 182 pathways present in all cancer types considered , a value that translated into 68 . 7% of pathways in GBMs , 56 . 5% in ovarian and 58 . 9% in breast cancers . Furthermore , we observed a small overlap of 98 IM-pathway pairs that appeared in all cancer types considered ( Fig . S2B , Table S5 ) . Since we determined the impact of each interacting miR on the fit of each pathway's enrichment score , an IM may be important to more than one pathway and vice versa . In Fig . S3A , we observed a logarithmic decay in the frequency distribution of the number of pathways an IM targeted in all cancer types . In turn , the frequency distribution of the number of IMs a given pathway is significantly linked to decreased exponentially as well ( inset , Fig . S3B ) . Obtaining auxiliary cancer-related information , we collected 72 cancer-related miRs from the literature , consisting of overlapping sets of 35 onco- , 42 tumor suppressor- and 32 metastamiRs [6] , [9] , [17] , [18] , [19] , [20] , [21] ( Fig . S1 ) . Furthermore , we utilized the HMDD database [23] pooling experimental evidence that a miR was involved in given cancer types . We also determined differentially expressed miRs with a t-test ( FDR<0 . 01 ) [32] using miR expression profiles of glioblastomas , ovarian and breast cancer . In Table S6 we ordered IMs according to their corresponding number of pathways in each cancer type . Specifically , IMs that were linked to an increasing number of pathways seemed to be enriched with literature curated cancer miRs , tend to be differentially expressed and experimentally indicated in the given cancer types . On a more quantitative basis , we grouped IMs according to their number of pathways in a given cancer type . In groups of IMs that were linked to at least k pathways we determined the number of literature-curated miRs . In a null-model , we randomly picked sets of literature-curated miRs and determined their enrichment in each group as the ratio of the observed and expected numbers . Fig . S4A suggests that groups of IMs with increasing numbers of pathways tend to be enriched with cancer miRs in all given cancer types . Analogously , we determined the enrichment of differentially expressed miRs and observed that such groups of IMs were predominantly enriched with differentially expressed miRs as well ( inset , Fig . S4A ) . Similarly , we calculated the enrichment of differentially expressed pathways as a function of the number of IMs of a given pathway . Distributions in Fig . S4B suggested that pathways with an increasing number of IMs had a heightened tendency to be differentially expressed in all tumor types considered . Utilizing data from the HMDD database [23] we collected information about miRs that were experimentally found to play a role in more than 90 cancer types . Focusing on 25 cancer types with at least 25 different , implicated miRs ( including glioblastoma , ovarian and breast cancer ) we constructed a bipartite matrix , indicating if a given miR was experimentally reported in a certain cancer type . Ward-clustering such a binary matrix , we observed two large clusters of miRs ( Fig . 2 ) . Counting the number of different cancer types a miR was experimentally found in , we observed that such clusters consisted of the most frequently indicated miRs ( histogram , Fig . 2 ) . Therefore , we expected that such clusters may be enriched with IMs . Indeed , our separate sets of IMs in glioblastoma , ovarian and breast cancer overlapped well with this general pattern of miR involvement in different tumor types . Applying a hypergeometric test we further checked if IMs were enriched among miRs that appeared in at least 3 different cancer types . Indeed , 106 IMs in GBMs occurred in such a set of miRs ( P<10−5 ) , while we found 107 in ovarian ( P<10−4 ) and 100 in breast cancers ( P<10−4 ) . Focusing on our set of 99 overlapping , overall important miRs ( OIM ) in GBMs , ovarian and breast cancers we also observed a significant overlap of 74 miRs ( P<10−5 ) . Furthermore , literature curated cancer miRs were largely placed in previously mentioned clusters as well . In particular , 38 cancer miRs overlapped with our set of 99 OIMs ( P<10−10 ) , suggesting that OIMs may play a central role in different cancer types . Utilizing such an overlapping set of 99 OIMs , we focused on connections to differentially expressed pathways and found a total of 93 pathways in glioblastoma , 55 in ovarian and 87 in breast cancers . Mapping the corresponding links between OIMs and these pathways in glioblastoma we constructed a binary matrix . Ward clustering allowed us to obtain two large clusters of either up- or down-regulated pathways that strongly corresponded to two groups of largely down- or up-regulated , differentially expressed OIMs ( Fig . 3A ) . Down-regulated pathways mostly revolved around neurotransmitter specific pathways while up-regulated pathways covered prominent signaling , regulation and transcription functions ( see for an enlargement ) . As for ovarian ( Fig . S6A ) and breast cancers ( Fig . S7A ) , we obtained similar results . Notably , we only observed interactions between OIMs and up-regulated pathways in ovarian cancers that largely revolved around signaling and regulation functions . Using such pairs of OIMs and pathways in GBMs , we retrieved all interactions between OIMs and LEGs in the corresponding pathways that were placed in the previously found clusters . Merging gene and miR expression data , we calculated Pearson's correlation coefficients using gene and miR expression profiles in glioblastoma and non-tumor control samples . As a measure of the difference between expression correlation coefficients in the disease ( rD ) and non-tumor control cases ( rC ) we Fisher-transformed correlation coefficients into Z-scores and calculated the corresponding change in correlation , ΔZ . A negative/positive value of ΔZ indicates a loss/gain of correlation in the disease case . Focusing on interactions between OIMs and the corresponding LEGs of pathways in these clusters we observed bimodal distributions of ΔZs in glioblastoma ( Fig . 3B ) . Notably , interactions between OIMs and LEGs that corresponded to down-regulated pathways and predominantly up-regulated miRs were characterized by a peak at ΔZ = −1 . 0 , pointing to a loss of expression correlation . Focusing on miR/gene interactions in the cluster of up-regulated pathways and largely down-regulated miRs we observed a peak at ΔZ = +1 . 0 , pointing to a gain of correlation . Analogously , we obtained such distributions for pairs of OIMs and LEGs in ovarian ( Fig . S6B ) and breast cancer ( Fig . S7B ) . Focusing on GBMs , we mapped all interactions between OIMs and LEGs we found in the corresponding clusters if their correlation change was |ΔZ|>1 . 0 . As for the cluster that revolved around down-regulated pathways and up-regulated OIMs ( Fig . 3C ) , we observed many interactions between differentially expressed OIMs and ITPR1 ( inositol 1 , 4 , 5-trisphosphate receptor type 1 ) with losses of expression correlations . Overall important miRs mapped in this analysis included miR-34a , -27b , -128ab and -15b . Focusing on the cluster composed by down-regulated pathways and largely up-regulated OIMs ( Fig . 3D ) , we found miR-21 and let-7i in interactions with losses of expression correlation and miR-137 in interactions that gained expression correlation . We mapped miRs and associated pathways in ovarian ( Fig . S6C ) and breast cancers as well ( Fig . S7CD ) . While we found a strong presence of signaling , transcription and translation related pathways in ovarian cancers , we also observed pathways that revolved around transcription factor E2F and the SFRS1 protein . Focusing on a cluster of up-regulated pathways in breast cancers and largely up-regulated miRs ( Fig . S7C ) we found down-regulated AKT3 that was interacting with a couple of up-regulated miRs . These results are discussed below ( see Discussion ) .
Although a growing appreciation of the importance of miRs in cancers is emerging , much remains unknown about their regulatory impact . Current knowledge appears rather scattered , focusing on single interactions between miRs and target genes of interest in a given cancer type . Here , we chose a different approach by utilizing pairwise interactions between miRs and target genes to identify combinations of important miRs ( IM ) and pathways in a given cancer type . A major criterion that may influence our results is the accuracy of computational methods that predict interactions between miRs and the UTRs of genes . Since such computational approaches suffer from false positives , we chose results of three different algorithms and demanded that each interaction was at least predicted twice , potentially allowing us to limit spurious signals [34] . We modeled the expression change of pathways comparing sets of cancer to non-tumor control cases as a function of the number of interactions between leading edge genes that drive the expression of a given pathway and miRs . We stress our initial assumption that the mere number of targeted LEGs in a pathway is a reasonable proxy to model the expression change of pathways in a disease , therefore allowing us to capture tumor specific effects . Although our approach did not account for any expression levels of miRs in given tumor types , we assume that the expression change of pathways is not only a matter of leading edge genes but the binding miRs as well . As such , we modeled expression change as a skeleton of miR interactions . Since such links strongly influence the flow of molecular information , we conclude that the consideration of miRs expression putatively won't override results that were largely imposed by the underlying topology of miR interactions . Furthermore , such an approach allows us to determine combinations of important miRs that potentially influence such expression changes through their targeted LEGs in the given pathways . Utilizing data of diverse cancer types , such as glioblastomas , ovarian and breast cancers , we clearly observed largely overlapping sets of IMs that were predominantly linked to differentially expressed pathways . Confirming our initial hypotheses , IMs with many pathways were predominately enriched with literature-curated cancer miRs and differentially expressed miRs . Besides , such pathway specific connections may be harnessed to predict meaningful sets of miRs that play a role in the underlying cancers . Notably , overall important miRs ( OIM ) in all cancer types coincided well with the most frequently indicated cancer –related miRs in different cancer types , indicating the relevance of our predictions . While the consideration of miR expression levels may change the number of IMs , such observations strongly suggest that a diminished set of OIMs will continue to show similar characteristics . Focusing on specific details of glioblastomas , ovarian and breast cancers , such cancer types are typically stratified by certain subtypes as indicated by subtle changes in gene expression profiles . While we acknowledge that pairs of pathways and important miRs may vary , we don't expect that the sets of IMs will dramatically change: considering that completely different cancer types with significant differences in their gene expression profiles provided largely overlapping sets of IMs , we expect that results that account for subtype information will be largely robust . Focusing on our set of 99 OIMs , we identified all interactions to LEGs in differentially expressed pathways . Comparing non-tumor control to disease cases , such interactions suffered partially from a massive loss of ( anti- ) correlation that were indicated by multimodal distributions of expression correlation changes . Dramatic changes of the expression correlation of interactions may therefore be considered to significantly influence the expression of LEGs , contributing to the perturbation of pathways in the underlying cancer types . As for qualitative observations of such OIM-LEG pairs we found that many differentially expressed miRs appeared interacting with ITPR1 in GBMs ( Fig . 3C ) . This receptor1 is central to many signaling GBM-relevant pathways , including NGF and Plc-γ1 signaling pathways as well as insulin regulation and diabetes related pathways . miR-34a has been found to play an important role in glioblastoma as a tumor suppressor [16] , [35] while being a mediator of p53 [14] , [36] , [37] , [38] , [39] in an interaction with a loss of expression correlation . Important targets of miR-34a included members of the Notch family and the oncogene c-met [40] . Specifically , we found an association of miR-34a with phospholipase C ( PLCB1 ) , which has recently been identified as a regulator of glioma cell migration [41] . The result of miR-27b was rather unexpected , since this miR has been reported up-regulated in gliomas [42] . However , the observed discrepancy may result from the experimental setup where the up-regulated miR-27b might have resulted from an inflammatory reaction [43] and originated from other than the glioma cells . Moreover , miR-27b has been identified as a pro-angiogenic miR in endothelial cells [44] and found to be involved in tumor angiogenesis [45] . Regarding the up-regulation of miR-27b in glioma cells , cell culture conditions used in [42] promote cell differentiation ( medium containing fetal bovine serum ) that may artificially affect the miRs expression profile . Therefore , we believe that the down-regulation of miR27b and its effects on calcium metabolism ( CALM3 , CACNB2 ) and exocytosis-related ( SNAP25 ) genes reflect the actual situation in GBMs . The down-regulation of miR-128ab in human glioma and glioblastoma cell lines has previously been reported [46] to increase the expression of ARP5 , Bmi-1 and E2F-3a , promoting neural stem cells renewal and regulate cell-cycle progression [46] . Beside miR-128ab being important regulators of brain cell proliferation , we indicated that miR128ab may also affect expression of genes involved in energy metabolism ( PFKM ) and transmembrane signal transduction ( SYT1 , EPB41 , ADCY3 ) . miR-15b has been identified as an inhibitor of glioma growth while cyclin E1 has been found as a target of miR-15b , suggesting its role in cell cycle regulation [47] . Here , we observed that serotonin receptor 4 ( HTR4 ) was down-regulated in glioblastoma samples , a process that is associated with up-regulation of miR-15b . The cluster composed of down-regulated pathways and largely up-regulated OIMs ( Fig . 3D ) revealed miR-21 , let-7i , and miR-137 to be involved in interactions with losses and gains of expression correlation , respectively . Putatively , miR-21 works as an ‘oncomiR’ , decreasing apoptosis in malignant cells while down-regulated miR-137 is involved in the differentiation of glioma stem cells [48] . Implicated in the development of glioblastomas [49] , [50] , knockdown of miR-21 leads to reduced cell proliferation , invasiveness , tumorigenicity and increased apoptosis [49] , [50] , [51] . Furthermore , miR-21 was reported to be involved in at least three tumor-suppressive pathways including mitochondrial apoptosis , p53 and TGF-β [50] , [52] , [53] , [54] pathways . Our results revealed further cancer-relevant target genes including STAG2 , CNOT6 , SOX2 , CDC25A and SFRS3 ( Fig . 3D ) . Specifically , STAG2 encodes a subunit of cohesion , a multimeric protein complex required for cohesion of sister chromatids after DNA replication . Furthermore , STAG2 is cleaved at the metaphase-to-anaphase transition to enable chromosome segregation [55] , [56] , [57] . Chromosomal instability , which leads to aneuploidy , loss of heterozygosity , translocations and other chromosomal aberrations is one of the hallmarks of cancer [57] . Robust STAG2 expression has been shown in non-neoplastic tissues while significant fractions of glioblastomas had completely lost expression of STAG2 [58] , suggesting that miR-21 may have both oncogenic and tumor-suppressive effects . A link between miR-21 and the p53 pathway could be CNOT6 ( Ccr4a ) , a deadenylase subunit of the Ccr4-Not complex that is involved in mRNA degradation [59] . Ccr4a , together with Ccr4b , has been identified as a key regulator of insulin-like growth factor-binding protein 5 , mediating cell cycle arrest and senescence through the p53-dependent pathway [60] , [61] . Moreover , CNOT6 plays an important role in chemotherapy resistance to cisplatin through down-regulation of DNA-damage response by targeting Chk2 [62] . miR-21 expression was shown up-regulated in response to ionizing radiation while the inhibition of miR-21 enhanced the radiation-induced glioblastoma cell growth arrest and increased the level of apoptosis . While this effect may be mediated by CDC25A [63] , our results suggested that CDC25A was targeted by miR-21 ( Fig . 3D ) . Additionally , Cdc25A appears to be a promising therapeutic target in glioblastomas as its levels were reported to correlate with Ki-67 labeling index [64] . Another target gene that we identified to be controlled by miR-21 , SFRS3 , is a pro-oncogene involved in mRNA and rRNA processing . Furthermore , SFRS3 has been reported as a critical factor for tumor induction , progression and maintenance [65] , [66] . Lastly , the association of miR-21 with SOX2 , a marker for undifferentiated and proliferating cells with up-regulated expression in glioblastomas [67] further underlined the importance of miR-21 for the pathogenesis of these tumors . Let-7 appears to be a tumor suppressor while inhibiting K-ras and C-myc [68] , [69] . In glioblastomas , overexpression of let-7 has been shown to decrease cell proliferation [70] . We found a link between let-7i and integrin β3 ( ITGB3 ) whose pro-apoptotic role has been reported in glioma cells [71] . miR-137 is also a putative tumor suppressor and is down-regulated in gliomas through a DNA hypermethylation mechanism [48] . Cooperating with miR-124 , miR-137 may suppress expression of phosphorylated Rb and CDK6 while inducing cell cycle arrest at G0/G1 in glioma cells [48] . Our results further suggested glioma relevant targets that are involved in AKT-mTOR signaling ( MAPKAPK2 and YBX1 ) ( Fig . 3D ) . The significance of other associated partners such as genes that encode ribosomal proteins RPL28 and RPS13 remains to be established . Mapping OIMs and their pathways in ovarian cancer revealed interactions between several miRs and transcription factor E2F and particularly between E2F3 and miRs-148b , -124 and -34a ( Fig . S6C ) . Indeed , miR-34a was shown to epigenetically govern the expression of E2F3 through methylation of its promoter [72] . In our analysis , miR-132 and miR-212 gain expression correlation in interactions with SFRS1 , a proto-oncogene that is involved in pre-mRNA splicing with the ability to change the splicing patterns of crucial cell cycle regulators and suppressor genes . Of particular interest is the observation that SFRS1 is up-regulated in many cancer types and therefore a potential target for cancer therapy [73] . Importantly , the role of these miRs and their interactions with target genes in ovarian cancers is not well understood . However , indications exist that both miRs that share a seed sequence may play a role since both miRs were found to be down-regulated by promoter methylation that contributes to pancreatic cancers [74] . The down-regulation of AKT3 upon interaction with several up-regulated miRs was the highlight observation in the cluster of up-regulated pathways in breast cancers ( Fig . S7D ) . AKT kinases are regulators of cell signaling in response to insulin and growth factors and are involved in a wide variety of biological processes including cell proliferation , differentiation , apoptosis , tumorigenesis as well as glycogen synthesis and glucose uptake . In our analysis , we found that AKT3 interacted with miRs-181ac , gaining expression correlation , while miR15a , -16 and -20a lost expression correlations with their target genes . In particular , miR-15a and -16 were already indicated as relevant in different cancers [75] . Furthermore , members of the miR-181 family were shown to induce sphere formation in breast cancer cells [76] .
|
We assume that a network of physical interactions between miRs and genes allows us to determine miRs that influence the expression of whole pathways in different tumor types . Specifically , we represented each pathway by an enrichment score and an array of miRs counting the number of genes in the pathway a given miR can bind . Despite the different nature of the considered tumor types , we obtained a large set of overlapping miRs using a machine-learning algorithm . Such associated miRs were enriched with literature-curated cancer and differentially expressed miRs and also coincided well with clusters of miRs that were experimentally indicated in numerous other cancer types . Focusing on such sets of miRs we observed that interactions with genes in differentially expressed pathways were characterized by massive gains/losses of expression correlations . Such drastic changes of miR and gene expression correlation indicate miR/gene pairs that may play a causal role in the underlying cancers .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"systems",
"biology",
"regulatory",
"networks",
"biology",
"computational",
"biology"
] |
2013
|
Important miRs of Pathways in Different Tumor Types
|
Spontaneous emergence of synchronized population activity is a characteristic feature of developing brain circuits . Recent experiments in the developing neo-cortex showed the existence of driver cells able to impact the synchronization dynamics when single-handedly stimulated . We have developed a spiking network model capable to reproduce the experimental results , thus identifying two classes of driver cells: functional hubs and low functionally connected ( LC ) neurons . The functional hubs arranged in a clique orchestrated the synchronization build-up , while the LC drivers were lately or not at all recruited in the synchronization process . Notwithstanding , they were able to alter the network state when stimulated by modifying the temporal activation of the functional clique or even its composition . LC drivers can lead either to higher population synchrony or even to the arrest of population dynamics , upon stimulation . Noticeably , some LC driver can display both effects depending on the received stimulus . We show that in the model the presence of inhibitory neurons together with the assumption that younger cells are more excitable and less connected is crucial for the emergence of LC drivers . These results provide a further understanding of the structural-functional mechanisms underlying synchronized firings in developing circuits possibly related to the coordinated activity of cell assemblies in the adult brain .
Coordinated neuronal activity is critical for a proper development and later supports sensory processing , learning and cognition in the mature brain . Coordinated activity represents also an important biomarker of pathological brain states such as epilepsy [1] . It is therefore essential to understand the circuit mechanisms by which neuronal activity becomes coordinated at a population level . A series of experimental results indicates that non-random features are clearly expressed in cortical networks [2–4] , in particular neuronal sub-networks , termed cliques [5] , have been shown to play a fundamental role for the network activity and coding both in experiments [6–10] as well as in models [11–14] . The identification of these small highly active assemblies in the hippocampus [6] and in the cortex [7–9] poses the question if these small neuronal groups or even single neurons can indeed control the neural activity at a mesoscopic level . Interestingly , it has been shown that the stimulation of single neurons can affect population activity in vitro as well as in vivo [15–25] . The direct impact of single neurons on network and behavioral outputs demonstrates the importance of the specific structural and functional organization of the underlying circuitry . Neurons having such a network impact were recently termed operational hubs [26] or driver cells [24] . It is thus critical to understand how specific network structures can empower single driver cells to govern network dynamics . This issue has been addressed experimentally in some cases . More specifically , in the developing CA3 region of the hippocampus , single GABAergic hub neurons with an early birthdate were shown to coordinate neuronal activity . These cells have a high functional connectivity degree , reflecting mainly the fact that they are activated at the onset of Giant Depolarizing Potentials ( GDPs ) , as well a high effective connectivity degree [17] . This therefore represents a simple case where the circuit mechanism , promoting a cell to the role of hub , is due to their exceptional number of anatomical links . But the picture can be quite different in other brain regions , as recently demonstrated in the developing Entorhinal Cortex ( EC ) [24] , where the driver cell population comprises both cells with a high functional out-degree , as well as low functionally connected ( LC ) cells . In order to understand the circuit mechanisms by which even a LC cell can influence population bursts we have upgraded and modified a network model based on excitatory leaky integrate-and-fire ( LIF ) neurons [12] , previously developed to reproduce the functional properties of hub neurons in the developing hippocampal CA3 area [17] . In such a model the population bursts ( PBs ) , corresponding to GDPs in neonatal hippocampus [27] , were controlled by the sequential and coordinated activation of few functional hubs . Notably , the perturbation of one of these neurons strongly impacted the collective dynamics and brought even to the complete arrest of the bursting activity , similarly to what experimentally found for the developing hippocampus in [17] . The model described in this paper contains two main differences with respect to the hippocampal model [12] . Firstly , it comprises both inhibitory and excitatory neurons , to account for the fact that , even though GABA acts as an excitatory neurotransmitter at early postnatal stages , some more developed neurons have already made the switch to an inhibitory transmission at the end of the first postnatal week in mice ( P8 ) , where most experimental data was obtained [28–30] . In that respect , the maturation profile of the entorhinal-hippocampal circuit was recently analyzed and it was shown that stellate cells in layer 2 of the medial EC were the first to mature , followed , in chronological order , by pyramidal cells in MEC layer 2 and CA3 [31] . This indicates a higher probability to find inhibitory GABAergic synapses in EC with respect to CA3 and it justifies our choice to include some inhibitory neuron in the model . Secondly , the developmental profile of the network is regulated only by the correlation between neuronal excitability and connectivity , while in [12] a further correlation was present . The anti-correlation between intrinsic excitability and the synaptic connectivity reproduces to some extent the homeostatic regulation of the intrinsic excitability described during neural development [32] . This model nicely mimicked the experimental observations in the EC similarly displaying the presence of driver cells with both low and high functional connectivity . The paper is organized as follows . Firstly , we will report experimental evidences of how the stimulation of single LC drivers can impact network synchonization in the developing EC . Secondly , we will show that our simple model can nicely reproduce such results . This to validate its applicability in order to understand the synchronization dynamics of the EC at the early stage of the development . Then , we will present a full characterization of the numerical model leading to a complete understanding of the mechanism underlying the PB generation and the impact of driver LC cells on population dynamics . We will conclude with a discussion of the obtained results and of their possible relevance in the context of neuroscience .
The main experimental observation at the rationale of this work is the existence of driver cells ( or Operational Hubs [26] ) in the mice EC during developmental stage [24] . Driver cells have been identified using calcium imaging experiments and they were characterized by the capability to impact network synchronization ( namely , GDPs′ occurence ) when externally activated/stimulated through intra-cellular current injection . Two classes of driver cells were identified: ( i ) those with high directed functional connectivity out-degree , early activated and playing a critical role in the network synchronizations ( driver hub cells ) and ( ii ) those recruited only in the later stages of the synchronization build up , which therefore are low functionally connected ( driver LC cells ) . The experimental setup used to identify , target and probe the single-handedly impact of neurons on spontaneous EC synchronization is schematized in Fig 1 ( a . E ) and S1 Fig . In brief , the functional connectivity of the cells has been measured during the spontaneous activity session , which preceded the single neurons’ stimulation session , both lasting two minutes . A directed functional connection from neuron A to B was established whenever the firing activity of A significantly preceded the one of neuron B ( more details can be found in Methods ) . The functional out-degree D i O of a neuron i corresponded to the percentage of imaged neurons which were reliably activated after its firing . Neurons in the 90% percentile of the connectivity distribution were classified as hub neurons early activated in the network synchronization . The protocol used for probing the impact of single neurons on the network dynamics was organized in three phases , each of two minutes duration: ( 1 ) a pre-stimulation resting period; ( 2 ) a stimulation period , during which a series of supra-threshold current pulses at a specific frequency νS have been injected into the cell; ( 3 ) a final recovery period , where the cell is no more stimulated . The frequencies νS employed to stimulate the single neurons have been selected to be of the order of magnitude of the average GDP frequency with the aim of revealing cell-network interaction . A cell was identified as a driver whenever the distributions of the Inter-GDP-Intervals ( IGIs ) were significantly different in the stimulation period with respect to the pre-stimulation and recovery periods ( see Methods for more details ) . A further indicator that we use to visualize the effect of the stimulation on the GDP deliver is the shift of the IGI phase Φ measured with respect to the pre-stimulation phase ( for the definition see Methods Eq ( 1 ) and in [33] ) . At a population level the stimulation may have an inhibitory ( excitatory ) effect corresponding to a slow-down ( acceleration ) of the GDP frequency associated with an increase ( decrease ) of the measured IGI corresponding to a positive ( negative ) phase shift . Two examples of driver LC cells , with D0 ≃ 7 − 8% , are reported in Fig 1 in the panels ( b-d . E ) and ( e-g . E ) . In the first case , upon stimulation the network dynamics accelerated , as testified by the decrease of the average IGI ( Fig 1 ( b . E ) ) and by the negative instantaneous phase shift of GDPs ( Fig 1d . E ) . In the second case , the stimulation led to a pronounced slow down of the average network activity ( as shown in Fig 1 ( e . E ) ) together with an increase of the instantaneous phase with respect to control conditions ( Fig 1 ( g . E ) ) . In both cases the removal of the stimulation led to a recovery of the dynamics similar to the control ones . A further extreme case of a silent cell , i . e . not spontaneously active and therefore with a zero ( out-degree ) functional connectivity , is shown in Fig 2 . This cell , when stimulated with different stimulation frequencies νS , revealed opposite effects on the network behaviour . At lower stimulation frequency ( νS = 0 . 33 Hz ) the cell activity induced an acceleration of the population dynamics ( see Fig 2 ( a ) –2 ( c . E ) ) , while at higher stimulation frequency ( νS = 1 Hz ) of the same neuron we observed a slowing down of the network dynamics ( see Fig 2 ( d ) –2 ( f . E ) ) . By considering a much larger pool of driver cells we have verified that the value of νS does not induce a systematic trend towards a slow down or an acceleration of the GDPs ( for more details see S2 Text and S3 Fig ) . We also tested the possibility that single neuron stimulation could modify other features of network synchronization besides the GDP frequency , in particular we focused on the width of the GDPs , i . e . on their time duration as defined in S1 Text . However , as shown in S2 Fig we could not find any statistical differences between stimulation epoch and pre/post stimulation periods ( see S1 Text for details ) . This doesn’t mean that other more subtle features of the GDPs could not be impacted by manipulating single neurons , but probably these modifications could be hardly discernible in view of the limited time resolution associated to calcium imaging . In order to mimic the impact of single neurons on the collective dynamics of a neural circuit , we considered a directed random network made of N LIF neurons [34 , 35] composed of excitatory and inhibitory cells and with synapses regulated by short-term synaptic depression and facilitation , analogously to the model introduced by Tsodyks-Uziel-Markram ( TUM ) [36] . In particular , synaptic depression was present in all the connections , while facilitation only in the connections targeting inhibitory neurons . This in agreement with recent experimental investigations of Layer II of the medial enthorinal cortex reporting evidences of short-term synaptic depression among excitatory neurons ( namely , stellate and pyramidal cells ) as well as among fast-spiking interneurons and pyramidal cells and of short-term facilitation in the connections from stellate cells towards low-threshold-spiking interneurons [37] ( see Methods for more details ) . As shown in [36 , 38 , 39] , these networks exhibit a dynamical behavior characterized by an alternance of short periods of quasi-synchronous firing ( PBs ) and long time intervals of asynchronous firing , thus resembling cortical GDPs’ occurrence in early stage networks . Similarly to the modeling reported in [12] , we considered neuronal intrinsic excitabilities negatively correlated with the total connectivity ( in-degree plus out-degree ) ( for more details see Definition of the Model in Methods and S4 Fig ) . The introduction of these correlations was performed to mimic developing networks , where both mature and young neurons are present at the same time associated to a variability of the structural connectivities and of the intrinsic excitabilities . Experimental evidences point out that younger cells have a more pronounced excitability , most likely due to the fact that their GABAergic inputs are still excitatory [40–42] , while mature cells exhibit a higher number of synaptic inputs and they do receive inhibitory or shunting GABAergic inputs [17 , 43] . The presence of inhibition and facilitation are the major differences from the model developed in [12] to simulate the dynamics of hippocampal circuits in the early stage of development , justified by the possible presence of mature GABAergic cells in the network . Using this network model , we studied the effect of single neuron current injection Istim on network dynamics , thus altering the average firing frequency of the neuron during the stimulation time , similarly to what done in the experiments . In the numerical investigations , at variance with the experiments , the stimulation delivered to the neurons is an unique supra-threshold step of duration of 48 seconds . In Fig 1 two representative driver LC cells are reported for comparison with the experiments . The first cell ( panels ( b-d . S ) of Fig 1 ) was a silent neuron in control conditions ( therefore with DO = 0 ) , that once stimulated could enhance of ≃ 30% the PB emission , thus leading to a decrease of the instantaneous phase Φ with respect to control condition . Panels ( e-g . S ) refer to a second neuron characterized by a low functional output connectivity , namely DO = 3% , whose stimulation led to a depression in the PB frequency ( as shown in panels ( e . S ) and ( f . S ) ) joined to an increase of the instantaneous phase of the network events with respect to control conditions ( as shown in panel ( g . S ) ) . These results compare quite well with the experimental findings reported in the same figure . Furthermore , analogously to what found in the experiment , Fig 2 ( a ) –2 ( f . S ) shows a silent neuron in control condition that once stimulated could lead to both enhancement or depression of the population activity depending on the level of injected current during stimulation . A full characterization of the network model concerning the impact on the network dynamics of each single neuron stimulation in relation to neuronal type , current injected and functional connectivity is detailed below . In order to explore the full dynamical range associated to the impact of single neuron stimulation on the network dynamics , we examined the response of the model network to two types of single neuron perturbations , i . e . single neuron deletion ( SND ) and single neuron stimulation ( SNS ) by employing the protocols introduced in [12] . In particular , the SND experiment consisted in recording the activity of the network in a fixed time interval Δt = 84 s when the considered neuron was removed from the network itself . While , the stimulation of the single neuron ( SNS ) was performed with a step of DC current of amplitude Istim for a time window Δt = 84 s . The recording of the activity in control condition was lasting 84 s as well , in order to compare directly the number of observed PBs during control and perturbation period . In particular , we tested the response of the network to a broad range of stimulation amplitudes varying from 14 . 5 mV ( slightly below the firing threshold for an isolated neuron Vth = 15 mV , see Methods ) to 18 . 0 mV with a step of 0 . 015 mV , inducing in the stimulated neuron a maximal firing frequency of ≃70 Hz . Typically the stimulated neuron fired with a frequency much higher than the frequency of neurons under control conditions ( i . e . in absence of any perturbation ) . As an example , for a stimulation current Istim = 15 . 90 mV the targeted neuron fired at a frequency ν ≃ 32 − 36 Hz well above the average ( ≃3 Hz ) and the maximal ( 22 Hz ) frequency of all neurons in control conditions . The SND represented an extreme version of the SNS , where the neuronal removal corresponded to the injection of an hyperpolarizing current inhibiting the neurons from firing spontaneously or in response to any synaptic input . In both SNS and SND experiments the impact of single neuron perturbation on the collective dynamics was measured by the variation of the PB frequency relative to control conditions . In general , we have classified a neuron as a driver cell whenever upon stimulation it is able to modify the PB frequency of at least 50% with respect to control conditions . In the specific , in analogy to what done in [24] , for SNS experiments we considered both enhancement and decrease in the PB activity . On the other hand , SND allowed us to directly identify the driver neurons which are fundamental for the PB build-up . Therefore in this case we limited to consider those cells whose SND led to a population decrease of at least 50% . The choice of this threshold is based on the analysis of the distribution of the number of PBs delivered during SNS and SND experiments compared to the PB variability observed in control conditions . This was preliminarily measured by considering the average and the standard deviation of the number of population bursts obtained by considering for each network realization 100 different initial conditions over a time window of 84 s . As we were interested in strong impact on the network dynamics , we decided to consider as significant the variations of the population activity which were well beyond the statistical fluctuations in the population bursting measured by three standard deviations . The choice of 50% fulfilled this condition in all the analyzed networks . Fig 3 ( a ) and 3 ( b ) report a comparison of the impact of SND and SNS ( with representative injected current of 15 . 90 mV ) on the PB activity . The removal of any of the four neurons labeled as ih1 , eh1 , eh2 , eh3 was able to arrest completely the bursting dynamics within the considered time window , while in other two cases ( for neurons ih2 and eh4 ) the activity was reduced of 60% with respect to the one in control conditions . For clarity , the used labels i/e stand for inhibitory/excitatory and h for hub , as we will show later this is related to the functional role played by these cells . For all the other neurons , the SND manipulation induced a non relevant modification in the number of emitted PBs , within the variability of the bursting activity in control conditions ( Fig 3 ( a ) ) . The SNS confirmed that the neurons ih1 , eh1 , eh2 , eh3 were capable to arrest the collective dynamics . Neurons eh4 and ih2 poorly impacted PB dynamics for the reported injected current , although for different values of Istim they were able to strongly influence the network dynamics ( as shown in the subsection Tuning of PBs frequency upon hubs’ and driver LC cells’ stimulations ) . At variance from what found in a purely excitatory network [12] , the SNS revealed also the presence of other 18 driver cells not identified by the SND capable to impact the occurence of PBs in the network ( Fig 3 ( b ) ) . For an equivalent random network , without any imposed correlation , SNS or SND affected the dynamics in a neglibile way producing a maximal variation of the bursting activity of 25-30% with respect to the control conditions ( see S5 ( a ) and S5 ( b ) Fig ) . To summarize , the presence of correlations among the neuronal intrinsic excitabilities and the corresponding structural connectivities was crucial to render the network sensible to single neuron manipulation . Differently from purely excitatory networks where SNS and SND experiments gave similar results , the inclusion of inhibitory neurons in the network promoted a larger portion of neurons to the role of drivers , and their properties will be investigated in the following . The role played by the neurons in the simulated network was elucidated by performing a directed functional connectivity ( FC ) analysis . In the case of the spiking network model , in order to focus on the dynamics underlying the PB build-up , the FC analysis was based on the first spike fired by each neuron in correspondence of the PBs . An equivalent information was provided in the analysis of the EC by considering the calcium signal onset to calculate the directed functional connectivity . The six neurons playing a key role in the generation of the PBs ( eh1−4 , ih1−2 ) were characterized by high values of functional out-degree , namely with an average functional degree DO = 68% ± 8% , ranking them among the 16 neurons with the highest functional degree . Given the high functional out-degree and their fundamental role in the generation of the PBs ( as shown by the SND in Fig 3 ( a ) ) , we identified these neurons as driver hub cells . The high value of DO reflected their early activation in the PB , thus preceding the activation of the majority of the other neurons . Next , we examined the structural degree of the neurons , specifically we considered the total structural degree KT , which is the sum of the in-degree and out-degree of the considered cell . As shown in Fig 3 ( f ) , we observed an anti-correlation among DO and KT where neurons with high functional connectivity are typically less structurally connected than LC neurons . This was particularly true for the six driver hubs , previously examined , since they were characterized by an average KT = 15 ± 3 , well below the average structural connectivity of the neurons in the network ( ≃ 20 ) . Concerning the excitability , the six driver hubs despite being in proximity of the firing threshold ( slightly above or below ) as shown in S6 ( a ) Fig , they were among the 25% fastest spiking neurons in control condition , ( as shown in Fig 3 ( c ) ) . In particular , the three neurons eh1 , eh2 , ih2 were supra-threshold , while neurons eh3 , eh4 , ih1 were slightly below the threshold . When embedded in the network their firing activity was modified , in particular three couples of neurons with similar firing rates can be identified , namely ( eh1 , ih1 ) , ( ih2 , eh2 ) and ( eh3 , eh4 ) , as reported in Table 1 . The direct structural connections present among these couples ( see also Fig 3 ( g ) ) could explain the observed firing entrainments , as discussed in details in the next subsection . When compared to the other hub neurons , the much lower activity of ( eh3 , eh4 ) , corresponding to twice the average frequency of the PBs in control condition , was related to the fact that these two neurons fired only in correspondence of the ignition of collective events like PBs and aborted bursts ( ABs ) , the latter being associated to an enhancement of the network activity but well below the threshold we fixed to detect PBs . This will become evident from the discussion reported in the subsection Synaptic resources and population bursts . As already mentioned , besides the six driver hubs , the SNS experiments revealed the existence of a different set of 18 drivers , whose activation also impacted the population dynamics , although they had no influence when removed from the network and therefore they were not relevant for the PBs build up . These neurons represented in Fig 3 with squares were characterized by a low FC , namely D0 = 13% ± 15% . Therefore , we have termed them driver LC cells representing the ones which reproduced the behaviour of the driver LC cells identified in the EC ( see Figs 1 and 2 and reference [24] ) . In the following we will refer to them as el… or il1 according to the fact that they are excitatory or inhibitory neurons , respectively ( note that only one LC driver was inhibitory ) . As shown in Fig 3 ( c ) , LC drivers were not particularly active ( with firing frequencies below 1 Hz in control conditions ) and in some cases they were even silent . Notably , under current stimulation they could in several cases arrest PBs or strongly reduce/increase the activity with respect to control conditions as shown in Fig 3 ( b ) for a specific level of current injection and also as discussed in detail in the following sections . Compared to the driver hubs , driver LC cells had a lower degree of excitability ( essentially they were all sub-threshold , see S6 ( a ) Fig ) , which resulted in a later recruitment in the synchronization build up , and as a consequence in a lower functional out-degree . Therefore , driver LC cells were not necessary for the generation of the PBs , playing the role of followers in the spontaneous network synchronizations . As shown in Fig 3 ( f ) , driver LC neurons were charaterized by a higher structural connectivity degree KT with respect to driver hubs , namely KT = 23 ± 3 , and the most part of them were structurally targeting the driver hubs either directly ( i . e . path length one ) or via a LC driver ( i . e . path length two , centered on a LC driver ) . In Fig 3 ( f ) , the two groups of drivers , hubs and LC cells , can be easily identified as two disjoint groups in the plane ( KT , D0 ) . These results indicated that driver hubs are not structural hubs , while the low functional connectivity neurons are promoted to their role of drivers due to their structural connections . This latter aspect will be exhaustively addressed in subsection Tuning of PBs frequency upon hubs’ and LC cells’s stimulation . In order to deepen the temporal relationship among neural firings leading to a PB , we examined the spikes emitted in a time window of 70 ms preceding the peak of synchronous activation ( see Methods for details ) . The cross correlations between the timing of the first spike emitted by each driver hub neuron during the PB build up are shown in S7 Fig ( Upper Sequence of Panels ) . The cross correlation analysis demonstrated that the sequence of activation of the neurons was eh1 → ih1 → ih2 → eh2 → eh3 → eh4 . The labeling previously assigned to these neurons reflected such an order . A common characteristic of these cells was that they had a really low functional in-degree DI as reported in Table 1 indicating that they were among the first to fire during the PB build-up . In particular , eh1 had a functional in-degree DI zero , revealing that it was indeed the firing of this neuron to initialize all the bursts and therefore it could be considered as the leader of the clique . A detailed inspection of the firing times , going beyond the first spike event , revealed the existence of more than one firing sequence leading to the collective neuronal activation: i . e . the existence of different routes to PBs . This is at variance with what found in [12] for a purely excitatory network , where only one route was present and all the PBs were preceded by the same ordered sequential activation of the most critical neurons . In particular , the neuron eh1 fired twice before the PBs ( see Fig 3 ( e ) ) , usually in-between the firing of eh2 and that of the pair ( eh3 , eh4 ) , and this represents the main route , occurring for ≃ 85% of the PBs . Along the second route ( present only for the ≃ 7% of the PBs ) , eh1 was firing the second time at the end of the sequence . The neuron eh1 fired essentially by following its natural period T 1 = τ m ln [ ( I e h 1 b - v r ) / ( I e h 1 b - V t h ) ] = 52 . 15 ms , and its second occurrence in the firing sequence depended on the delay among the firing of the other neurons . As a matter of fact we verified that the elimination of the second spike emitted by eh1 from the network dynamics didn’t prevent , and didn’t delay , the onset of the PB and had only a marginal effect on the firing of a very limited number of neurons in the PB . Therefore we can conclude that it is not essential to the PB build up . The two routes leading to the PB build-up are shown in Fig 3 ( e ) . To observe a PB the six driver hubs should fire not only in an ordered sequence , as shown in Fig 3 ( e ) , but also with defined time delays , their average values with the associated standard deviations are reported in S1 Table for the two principal routes . These results clearly indicate that the six driver hubs are arranged in a functional clique whose activation was crucial for the PB build-up . In the period between the occurence of two PBs , the driver hubs in the clique could be active , but in that case they did not show the precise sequential activation associated to the main and secondary route , see the out-of-burst results reported in the Lower Sequence of Panels in S7 Fig . A remarkable exception is represented by the case of the ABs , in that case PBs are not triggered despite the presence of the right temporal activation of all the hubs in the clique , due to the lack of synaptic resources ( as discussed in details in subsection Synaptic resources for population bursts ) . Out of PBs and ABs , we registered clear time-lagged correlations only for those neuronal pairs sharing direct structural connections ( shown in Fig 3 ( g ) ) : namely , eh1 → ih1 , ih2 → eh2 and eh2 → ( eh3 , eh4 ) . The firing delays of these neuronal pairs were not particularly altered also out of burst with respect to those measured during the burst build-up and reported in S1 Table . As shown in Fig 3 ( g ) , the eh3 neuron represented the cornerstone of the clique , receiving the inhibitory input coming from the structural pair ( eh1 , ih1 ) and the excitatory one from the pair ( ih2 , eh2 ) , with the activity of the neurons within each pair perfectly frequency locked . More specifically , eh1 entrained the activity of ih1 ( below threshold in isolation ) so that both neurons before a PB fired with a period quite similar to the natural period of eh1 . The other pair ( ih2 , eh2 ) was controlled by the inhibitory action of ih2 that slowed down the activity of eh2 , whose natural period was 60 . 6 ms , while before a PB ih2 and eh2 both fired with a slower period , namely 72 ± 2 ms . As it will be explained in details in the next two subsections , the two requirements to be fulfilled for the emergence of PBs are the availability of sufficient synaptic resources at neurons eh3 and eh4 and the coordinated activation of eh1 ( and ih1 ) with the pair ( ih2 , eh2 ) , in the absence of any synaptic connection between the two pairs . Next we analyzed the relation between the evolution of synaptic resources in the hub driver cells and the onset of the PB . The availability of synaptic resources was measured by the effective efferent synaptic strength XOUT as defined in Eq ( 8 ) . In particular , we considered the available resources only for the hub neurons eh3 and eh4 which were the last neurons of the clique to fire before the PB ignition . We have examined only these two hub neurons , because whenever eh3 and eh4 fired , a burst or an AB was always delivered . Neurons eh3 , eh4 were receiving high frequency excitatory inputs from eh2 ( although the natural firing of eh2 was slowed down by the incoming inhibition of ih2 ) and high frequency inhibitory inputs from ih1 ( entrained by the eh1 , the neuron with highest firing frequency in the network ) . This competitive synaptic inputs resulted in a rare activation of eh3 compared to the higher frequency of excitatory inputs arriving from eh2 . The period of occurrence of the ABs was comparable to the average interval between PBs ( namely , TPB = 1 . 4±1 . 0 s ) and ABs were preceded by the sequential activations of the six critical neurons of the clique in the correct order and with the required delays to ignite a PB . The number of observed ABs was 66% of the PBs , thus explaining why the average firing period of eh3 and eh4 was T e h 3 = 0 . 8s≃ T P B / ( 1 + 0 . 66 ) , since their firing always triggered a PB or an AB . To understand why in the case of ABs the sequential activation of the neurons of the clique did not lead to a PB ignition , we examined the value of synaptic resources for regular and aborted bursts , as shown in Fig 4 ( a ) . From the figure it is clear that X e h 3 O U T and X e h 4 O U T should reach a sufficiently high value in order to observe a PB , otherwise one had an AB . Furthermore , the value reached by X e h 3 O U T and X e h 4 O U T was related to the time passed from the last collective event and thus the requirement of a minimal value of the synaptic resources to observe a PB set a minimal value for the IGI , i . e . the interval between two PBs . As a matter of fact , as shown in Fig 4 ( b ) the IGI values grew almost linearly with the values reached by X e h 3 O U T just before the PB , at least for X e h 3 O U T < 0 . 9 . At larger values the relationship was no more linear and a saturation was observable , due to the fact that X e h 3 O U T could not overcome one . We could conclude that the slow firing of the couple ( eh3 , eh4 ) , moderate by the inhibitory action of ih1 on eh3 , was essential to ignite a PB , since a faster activity would not leave to the synapses the time to reach the minimal value required for a PB ignition , namely X e h 3 out * = 0 . 793 and X e h 4 out * = 0 . 666 . This could be better understood by reconsidering the SND experiment on ih1 , as expected the resection of neuron ih1 from the network led to a much higher activity of neurons eh3 and eh4 , as shown in S8 Fig . However , this was not leading to the emission of any PBs , because in this case the value of X e h 3 O U T and X e h 4 O U T remained always well below the value required for a PB ignition . Moreover , also among the network synchronization events classified as PBs we observed a variability in the number of neurons taking part to a PB . A detailed analysis reported in S3 Text reveals that at least two kind of events can be identified: one which sees the contribution of almost all the active neurons and another one where the participation is more limited ( see S9 Fig ) . These results resemble the ones reported experimentally for the EC in [24] . Furthermore , the analysis shows a clear correlation between the available synaptic resources of the drivers controlling the population activity and the entity of the observed PBs . In order to better understand the role played by the hub and LC drivers for the collective dynamics of the network , we performed SNS experiments for a wide range of stimulation currents . The results of this analysis for currents in the range 14 . 5 mV ≤ Istim ≤ 18 mV are shown in Fig 5 ( where all the driver hubs and six representative cases of driver LC cells are reported ) and in Fig 6 ( a ) . The driver hub neurons could , upon SNS , usually lead to a reduction , or silencing , of PBs , apart for two cells ( namely , eh1 and eh4 ) which , for specific stimulation currents , could even enhance the population activity . On the other hand , the 18 driver LC cells can be divided in two classes LC1 and LC2 according to their influence on the network dynamics upon SNS: a first group of 14 driver LC1 cells able mainly to reduce/stop the collective activity , and in few cases to increase the PB frequency , and a group of 4 LC2 neurons capable only to enhance the PB frequency . The three neurons el1 , el2 and el3 , previously considered in subsection Numerical evidences of driver LC cells for comparison with experimentally identified LC cells , belonged to the class LC1 ( see Figs 1 and 2 ) , while we have no experimental examples of LC2 cells . For what concerns the driver hubs’ dynamics , PBs were generated in the network whenever the hubs eh2 and ih1 , both structurally connected to eh3 , were stimulated with currents smaller than the excitability I e h 1 b of the leader of the clique and within a specific interval ( see Fig 5 ( b ) and 5 ( e ) ) . This means that in order to have a PB both neurons controlling eh3 should not fire faster than the leader of the clique . If this was not the case , the inhibition ( originating from ih1 ) would not be anymore sufficient to balance the excitation ( carried by eh2 ) or viceversa , thus leading eh3 to operate outside the narrow current window where it should be located to promote collective activity ( see Fig 5 ( c ) ) . In the case of ih2 and eh4 the SNS produced a less pronounced impact on the PB activity , their stimulation could never silence the network ( as shown in Fig 5 ( d ) and 5 ( f ) ) , apart in two narrow stimulation windows for ih2 . This is in agreement with what reported in Fig 3 ( a ) for the SND , since the removal of neurons eh4 and ih2 only reduced the occurrence of PBs of ≃ 60% .
We have developed a simple brain circuit model to mimic recent experimental results obtained in cortical slices of the mice Entorinhal Cortex during developmental stages [24] . These analysis revealed the existence of high and low functionally connected driver cells able to control the network dynamics . The fact that functional hubs can orchestrate the network dynamics is somehow expected [17 , 44] , while the existence of driver neurons with low functional out-degrees has been revealed for the first time in [24] . In this paper , we focused mainly on the analysis of these latter class of drivers . which in control conditions were essentially irrelevant for the build-up of the GDPs . On the contrary , if single-handedly stimulated they could nevertheless strongly modify the frequency of occurrence of GDPs , as evident from the experimental findings reported in Figs 1 ( b ) –1 ( g . E ) and 2 ( a ) –2 ( f . E ) . In particular , their stimulation could lead both to an enhancement as well as to a reduction of the population activity ( GDPs’ frequency ) . Quite remarkably , some of the driver LC cell were able to perform both these tasks as an effect of different stimulation frequencies as revealed by the experiment shown in Fig 2 ( a ) –2 ( f . E ) . Furthermore we have demonstrated that the experimental findings could be replicated in a simple spiking neural network model made of excitatory and inhibitory neurons with short-term synaptic plasticity and developmentally inspired correlations ( see Figs 1 ( b ) –1 ( g . S ) and 2 ( a ) –2 ( f . S ) ) . The analysis of the model has allowed to understand the fundamental mechanisms able to promote a single neuron to the role of network driver without being a functional hub , as usually expected . In the model , all the driver neurons able to influence the network dynamics could be identified and they could be distinguished in neuronal hubs characterized by high out-degree or low functionally connected drivers . Functional hubs are highly excitable excitatory and inhibitory neurons arranged in a clique , whose sequential activation triggered the Population Bursts ( analogous to GDPs ) . This in agreement with recent experimental evidences that small neuronal groups of highly active neurons can impact and control cortical dynamics [7–10] . On the other hand , driver LC cells are characterized by a lower level of excitability , but a higher structural connectivity with respect to driver hubs . Due to their low activity and functional connectivity in control conditions , these neurons were not fundamental for the PBs development , but were passively recruited during the burst , or even completely silent . The LC drivers can be divided in two classes LC1 and LC2 according to their influence on the population activity whenever stimulated with different values of DC current: the majority of them were able both to enhance and reduce ( or even set to zero ) the frequency of occurrence of the PBs ( LC1 ) , while a small group was able only to enhance the PBs’ frequency with respect to control conditions ( LC2 ) . Noticeably , driver LC1 cells were structurally connected to the hubs ( directly or via a bridge LC cell ) . Therefore , whenever stimulated they can influence the network activity by acting on the clique dynamics . In most cases , even if these cells were excitatory , their action on the network was mainly depressive , since either they stimulated directly inhibitory hubs or the inhibitory LC1 driver , which acted as a bridge over the clique . In more than the 50% of the cases ( 8 over 14 ) whenever brought over threshold driver LC1 cells led to a complete arrest of the PB activity . Driver LC2 cells instead were silent in control conditions and highly structurally connected , therefore they were putative structural hubs . As a matter of fact , whenever brought supra threshold they favoured a more regular collective dynamics . The activation of the many efferent connections of LC2 drivers led to the creation of many alternative pathways for the PB igniton , in a sort of homeostatic regulation of the network which led to an optimal employ of the synaptic resources [45] with the corresponding disappearence of the aborted bursts , largely present in control conditions . Furthermore , we have shown that the stimulation of single driver LC cells was not only able to alter the collective activity but also to deeply modify the role of neurons in the network , such that some neurons can be promoted to the role of driver hubs or driver hubs can even lose their role ( see Fig 6 ( b ) and 6 ( c ) ) . At variance with purely excitatory networks [12] , the synchronized dynamics of the present network , composed of excitatory and inhibitory neurons , is less vulnerable to targeted attacks to the hubs [46 , 47] . As demonstrated by the fact that different firing sequences of hub neurons can lead to population burst ignitions ( see Fig 3 ( e ) ) and that hubs can be easily substituted in their role by driver LC cells when properly stimulated . The robustness of the synchronized dynamics is confirmed by the fact that the presence of channel noise , up to quite large noise strength , does not substantially modify the composition neither of the functional clique nor of the LC drivers ( for more details on the analysis see S4 Text and S11 Fig ) . The experimental analysis of the population activity in the EC has revealed the existence of two kind of synchronization events: global , where most of the neurons were involved , and local , where only a sub-group took part to the burst [24] . In our model we have shown a similar effect that can be explained in terms of the available synaptic resources of the driver cells ( for more details see S3 Text and S9 Fig ) . It has been reported in different contexts [48 , 49] that the strength of the connectivity between neurons is a key variable defining neuronal ensemble dynamics . However , whether the mechanism that we have found to be at the basis of this variability could explain also the experimental findings in [24] is an interesting question that cannot be answered with our experimental set-up , and that we leave open for future investigation . Another relevant aspect is that the inclusion of inhibitory neurons in the network did not cause a trivial depressing action on the bursting activity , as it could be naively expected , but instead they can play an active role in the PB build-up . Our analysis clearly demonstrate that their presence among the driver cells is crucial in determining and controlling the PB activity , somehow similarly to what found in [20] where it has been shown that the emergence of sharp-wave in adult hippocampal slices was controlled by single perisomatic-targeting interneurons . We expect that the model described in this paper could also find application for other developing circuits , beyond EC , where inhibitory GABAergic synapses are also present . Our results suggest that inhibitory neurons can have a major role in information encoding by rendering on one side the population dynamics more robust to perturbations of input stimuli and on another side much richer in terms of possible repertoire of neuronal firings . These indications confirm the key role of inhibitory neurons in neural dynamics , already demonstrated for the generation of brain rhytms [50 , 51] and for attentional modulation [52] . Entorhinal cortex is involved in human mesial temporal lobe epilepsy [53] . A recent in vitro study on the onset of seizure-like events in EC slices [54] revealed fast-rising and sustained extracellular potassium increases associated to interneuronal network activity consistently preceding the initiation of seizures , supporting a key role of interneuron activity in the EC in focal seizure generation . There results are consistent with our observation that inhibitory neurons play a key role in the generation and orchestration of network synchronizations . The main ingredient introduced in our model to mimic the developmental stage was an anti-correlation between intrinsic excitability and the synaptic connectivity inspired by homeostatic regulation mechanisms observed during neural development [32] . Interestingly a similar phenomenon co-regulating neuronal connectivity and excitability could possibly occur in adult circuits due to functional switches or functional losses ( such as those following an injury ) . In such cases the prolonged absence of functional input in neurons could synergistically lead to synaptic depletion and increased neural excitability . Therefore , massive pre-synaptic neural loss in adults circuits could as a consequence enhance the neural excitability of the post-synaptic targets [9] , possibly upgrading the role of these cells to functional hubs with eventual pathological consequences as we could imagine in the case of epilepsy . Recently there has been a renovated interest on the existence and role of neuronal cliques within the brain circuitries [13 , 55] . Cliques have been proposed as structural functional multiscale computational units whose hierarchical organization can lead to increasingly complex and specialized brain functions and can ground memory formations [55] . In addition , activation of neuronal cliques as in response to external stimuli or feedforward excitation can lead to a cascade of neuronal network synchronizations with distinct spatio-temporal profiles [13] . Our results provide a further understanding on how cliques can emerge ( spontaneously during development ) and modify ( in response to stimuli similarly to the SNS here discussed ) with a consequent reshaping of the spatio-temporal profile of the dynamics of the network in which the clique is embedded . Notably , it is the presence of inhibitory neurons within the network to favour the emergence of different cliques by empowering drivers with different functional connectivity degree . While driver functional hubs guarantee the functioning of the network synchronization in absence of stimuli ( such as during development and in non-stimulated conditions ) , LC drivers widen the ability of the network to play distinct synchronization profiles ( i . e . spatio-temporal activations ) possibly underlying emergent functions within the brain networks . Finally , our results could be of some relevance also for the control of collective dynamics in complex networks [56 , 57] . Usually the controllability of complex networks is addressed with linear dynamics [58 , 59] . However , at present there is not a general framework to address controllability in nonlinear systems , in this context the SND and SNS protocols we developed for pulse-coupled networks could be extended to general complex networks as a tool to classify driver nodes and as a measure of controllability [60] . Control methods based on linear stability analysis for seizure propagation on large-scale brain networks have been recently reported [61] . Our methodology based on the SNS and SND protocols could be useful to go beyond this approach in order to single out the epileptogenic zone and the brain areas involved in the propagation of the seizure in patient specific brain models [62] .
|
There is timely interest on the impact of peculiar neurons ( driver cells ) and of small neuronal sub-networks ( cliques ) on operational brain dynamics . We first provide experimental data concerning the effect of stimulated driver cells on the bursting activity observable in the developing entorhinal cortex . Secondly , we develop a network model able to fully reproduce the experimental observations . Analogously to the experiments two types of driver cells can be identified: functional hubs and low functionally connected ( LC ) cells . We explain the role of hub neurons , arranged in a clique , for the orchestration of the bursting activity in control conditions . Furthermore , we report a new mechanism which can explain why and how LC drivers emerge in the structural-functional organization of the entorhinal cortex .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion"
] |
[
"action",
"potentials",
"medicine",
"and",
"health",
"sciences",
"neural",
"networks",
"nervous",
"system",
"population",
"dynamics",
"membrane",
"potential",
"electrophysiology",
"neuroscience",
"network",
"analysis",
"computational",
"neuroscience",
"population",
"biology",
"mood",
"disorders",
"computer",
"and",
"information",
"sciences",
"depression",
"animal",
"cells",
"mental",
"health",
"and",
"psychiatry",
"cellular",
"neuroscience",
"cell",
"biology",
"anatomy",
"synapses",
"single",
"neuron",
"function",
"neurons",
"physiology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"computational",
"biology",
"neurophysiology"
] |
2018
|
Modeling driver cells in developing neuronal networks
|
Risk factors associated with L . donovani visceral leishmaniasis ( VL; kala azar ) relapse are poorly characterized . We investigated patient characteristics and drug regimens associated with VL relapse using data from Médecins Sans Frontières - Holland ( MSF ) treatment centres in Southern Sudan . We used MSF operational data to investigate trends in VL relapse and associated risk factors . We obtained data for 8 , 800 primary VL and 621 relapse VL patients treated between 1999 and 2007 . Records of previous treatment for 166 VL relapse patients ( 26 . 7% ) were compared with 7 , 924 primary VL patients who had no record of subsequent relapse . Primary VL patients who relapsed had larger spleens on admission ( Hackett grade ≥3 vs0 , odds ratio ( OR ) for relapse = 3 . 62 ( 95% CI 1 . 08 , 12 . 12 ) ) and on discharge ( Hackett grade ≥3 vs 0 , OR = 5 . 50 ( 1 . 84 , 16 . 49 ) ) . Age , sex , malnutrition , mobility , and complications of treatment were not associated with risk of relapse , nor was there any trend over time . Treatment with 17-day sodium stibogluconate/paromomycin ( SSG/PM ) combination therapy vs 30-day SSG monotherapy was associated with increased risk of relapse ( OR = 2 . 08 ( 1 . 21 , 3 . 58 ) ) but reduced risk of death ( OR = 0 . 27 ( 0 . 20 , 0 . 37 ) ) , although these estimates are likely to be residually confounded . MSF operational data showed a crude upward trend in the proportion of VL relapse patients ( annual percentage change ( APC ) = 11 . 4% ( −3 . 4% , 28 . 5% ) ) and a downward trend in deaths ( APC = −18 . 1% ( −22 . 5% , −13 . 4% ) ) . Splenomegaly and 17-day SSG/PM vs 30-day SSG were associated with increased risk of VL relapse . The crude upward trend in VL relapses in Southern Sudan may be attributable to improved access to treatment and reduced mortality due to SSG/PM combination therapy .
Visceral leishmaniasis ( VL , kala-azar ) is a systemic parasitic disease caused by the Leishmania donovani species complex . VL manifests with irregular bouts of fever , substantial weight loss , hepatosplenomegaly , pancytopenia , and susceptibility to opportunistic infection [1] . VL is typically fatal unless treated . In immunocompetent individuals , effective drug treatment reduces Leishmania amastigotes to a level undetectable in aspirates . An effective life-long cellular immune response normally develops , and residual parasites are suppressed [2] . Despite apparent clinical and parasitologic response to treatment , a proportion of VL patients who are apparently otherwise immunocompetent , have a recurrence of VL . This usually occurs within 6 months of treatment [1] , [3] , and later recurrence is rare – suggesting that recrudescence rather than re-infection is the usual mechanism of relapse . While the role of HIV infection in VL relapse is well-documented , for example among patients in southern Europe , risk factors for VL relapse among HIV-negative patients in VL-endemic regions of Africa remain poorly characterised . Médecins Sans Frontières - Holland ( MSF ) has treated >80 , 000 VL patients in Sudan and Ethiopia since 1989 , and has maintained an electronic record of patient characteristics , treatments and outcomes . Although these routinely-collected data are intrinsically incomplete due to the very challenging environment in which they are collected , our previous analyses have yielded important findings [3]–[10] . Most recently , we analysed risk factors for VL relapse in patients in Northern Ethiopia who were co-infected with HIV [11] . In contrast , the prevalence of HIV in Southern Sudan has remained the lowest in Africa until very recently [12] . Here we investigate risk factors for VL relapse among patients from this largely HIV-negative population who were treated at MSF clinics in Southern Sudan from 1999 to 2007 .
Diagnostic , treatment , and discharge procedures used were consistent with WHO guidelines [13] . VL was diagnosed in clinical suspects by high titer ( ≥1∶6 , 400 ) antibodies to Leishmania ( freeze-dried Leishmania antigen supplied by the Royal Tropical Institute , Amsterdam , The Netherlands ) in a direct agglutination test ( DAT ) ; or by microscopy of splenic or lymph node aspirates; or ( since 2004 ) by rK39 rapid diagnostic test ( DiaMed-IT-Leish supplied by DiaMed AG , Cressier sur Morat , Switzerland ) ; or on rare occasions when laboratories were not functioning , by clinical judgment ( criteria: fever >2 weeks with exclusion of malaria and either splenomegaly or lymphadenopathy and wasting ) . Until 2002 , standard treatment for primary VL comprised daily IM injections of sodium stibogluconate ( SSG; Albert David Ltd , Calcutta; supplied by the International Dispensary Association , Amsterdam , The Netherlands ) at a dose of 20mg/kg/day ( minimum dose , 200mg; no maximum dose ) for 30 days . Between 2001 and 2003 , SSG monotherapy as standard treatment was replaced with combination therapy of paromomycin ( PM ) plus SSG . This comprised 17 daily intramuscular injections of SSG 20mg/kg and paromomycin sulfate ( Pharmamed Parenterals Ltd , Malta , supplied by the International Dispensary Association , Amsterdam , The Netherlands ) at a dose of 15mg/kg ( equivalent to ∼11mg/kg of PM base ) . SSG/PM was initially withheld from females of childbearing age , and since 2004 withheld only from pregnant women . SSG/PM was administered only at MSF treatment centres with a permanent expatriate doctor , and these centres also had access to parenteral antibiotics and therapeutic feeding . In 2003 MSF introduced a formal risk assessment using a scoring system based on known risk factors for adverse outcomes [3] . Thereafter , critically ill patients at high risk of death were treated initially in special care areas with rehydration , therapeutic nutrition , antibiotics , and liposomal amphotericin B ( AmBisome® , Gilead Pharmaceuticals ) ; when their condition improved , treatment continued with SSG/PM . Before the introduction of this “combined AmBisome-paromomycin-stibogluconate treatment” ( CAPST ) in 2003 , treatment with AmBisome was unusual . The prohibitive cost of AmBisome was reduced for public sector agencies in developing countries in 2007 , allowing all critically ill patients to be treated with a full course of AmBisome . Cases of primary VL who did not respond clinically to treatment underwent a test-of-cure aspirate from spleen or lymph node . If the test-of-cure result was positive , daily SSG injections were continued until two consecutive weekly test-of-cure results were negative . Patients whose aspirate remained positive after ≥60 SSG injections were treated with combinations of SSG , PM and AmBisome . On discharge , patients were given an identification card to be presented in case of re-admission for relapse or for post kala-azar dermal leishmaniasis ( PKDL ) . Patients with possible relapse of VL ( i . e . patients who presented their identification card from previous treatment or who verbally reported previous treatment ) were diagnosed by splenic or lymph node aspirate . Relapsed VL patients were treated with 17 days of SSG/PM plus further daily injections of SSG until two weekly test-of-cure results were negative . Patients who relapsed ≥2 times were treated with 6 IV doses of 4–6mg/kg of AmBisome on alternate days . In patients with relapses , tuberculosis ( TB ) and HIV were considered as possible co-factors and appropriate tests done . In rare cases of repeated VL relapses , empiric TB treatment was sometimes given . Patients with severe PKDL were treated with SSG or , if not responding , a combination of SSG/PM , until their condition improved . Since 1999 we have made an electronic archive of VL patient records from MSF treatment centres in Southern Sudan . Demographic , diagnostic , treatment , and discharge data were handwritten on a card during a patient's stay at a treatment centre . MSF staff at our Lokichoggio base on the Sudan/Kenya border entered data from these cards into EpiInfo , version 6 ( Centres for Disease Control and Prevention ) . Due to time and resource constraints , all data were single-entered . During the period 1999–2007 , MSF treatment centres were situated in Upper Nile State ( Abuong , Atar , Bimbim , Magang , Rupbuot , Wudier ) , Jonglei State ( Lankien , Pieri , Pading ) , and Unity State ( Bil , Leer , Nimne , Thonyor ) . For analysis of trends we used MSF operational data , which provided monthly summary statistics on total VL treatment activity ( admissions , diagnoses and outcomes ) at each treatment centre . Anomalous , inconsistent , or missing values were identified and corrected if possible , otherwise recoded as missing . Adult ( age >19 years ) body mass index ( BMI ) values were re-calculated from weight and height data . Weight-for-height ( WFH ) Z-scores for children age <5 years were generated using WHO Anthro 2005 software ( World Health Organisation , Geneva ) . BMI-for-age Z-scores for patients 5–19 years old were calculated using WHO Survey 2007 software ( World Health Organisation , Geneva ) [14] . This study was concerned only with patients with primary VL ( i . e . no record of previous treatment ) during 1999–2007 . In the first year ( 1999 ) , Lankien was the only functioning VL treatment centre in the area , with only 443 primary VL admissions and a recorded relapse rate of 3 . 7% by passive case-finding; we consider it unlikely that any of these primary VL patients were misdiagnosed VL relapse patients who had been treated elsewhere , because VL treatment had not been available elsewhere in the region . Patients who subsequently relapsed were identified by matching their primary and relapse VL treatment records , using their patient identifier , sex and age . Patients who died or defaulted ( self-discharge against medical advice ) during treatment were excluded from our analysis of risk factors for relapse . Our primary outcome was relapse after treatment for primary VL . We also analysed death during VL treatment as a secondary outcome in a comparison of SSG/PM vs SSG . We used the following admission data: age , sex , treatment centre , calendar year , self-reported duration of illness ( ‘time-to-presentation’ ) , walking status for patients ≥5 years old ( walking normally , with a stick , or with assistance or carried on stretcher - a marker for general weakness ) . Data on admission and discharge included: body mass index ( BMI ) ( for patients >19 years old ) , BMI-for-age Z-score ( for patients 6–19 years old ) [14] , weight-for-height ( WFH ) Z-score ( for patients ≤5 years old ) , spleen size ( Hackett grade ) [15] , and haemoglobin ( g/dl ) ( measured by color densitometry before 2006; Haemocue thereafter ) . Malnutrition in appropriate age groups was defined as: BMI<14kg/m2; BMI-for-age Z-score<−3; and WFH Z-score<−3 . Treatment data included: drug used and occurrence of diarrhoea , bleeding and vomiting . The majority of patients were treated at one centre ( Lankien ) which differed from the other ( smaller ) treatment centres in that expatriate and/or experienced national medical staff were stationed there permanently ( except when evacuated for reasons of security ) . Hence , treatment centre was coded dichotomously to indicate Lankien or ‘other’ treatment centre . We compared patients and outcomes by Chi-squared test and t-test . Odds ratios for VL relapse , adjusted a-priori for age , sex , treatment centre , and time ( calendar year ) , were estimated for all variables; those variables which were associated with VL relapse were then included , with the a-priori confounders , in a multivariable logistic regression model . All statistical analyses were performed using Stata Release 10 ( StataCorp . 2007 . Stata Statistical Software: Release 10 . College Station , TX ) . Trends based on MSF operational data ( summary statistics ) were analysed using US National Cancer Institute Join-point regression software , version 3 . 0 ( http://srab . cancer . gov/joinpoint/ ) , which performs linear regression to estimate the annual percent change ( APC ) in the dependent variable ( rates are assumed to change at a constant percentage of the rate of the previous year , ie . to change linearly on a log scale ) and the number and location of join-points ( points at which trends change ) [16] . The software performs pair-wise comparisons of models differing by one join-point to determine the model with the optimum fit to the data series . An overall significance level of 5% was adopted for the comparisons of models applied to each data series . We tested different permutations of the maximum number of join-points ( 2 , 3 or 4 ) , the minimum number of observations between joinpoints ( 3 or 4 ) and the minimum number of observations between joinpoints and the ends of the data ( 2 or 3 ) . ( 1 ) Patients without an enlarged spleen could not undergo spleen aspirate , and lymph node aspirates were performed by MSF at only one centre ( Lankien ) , hence we suspected that some VL patients diagnosed by DAT alone may have been misdiagnosed , because DAT has a false positive rate of ≥10% [17] , [18] . We therefore performed a sensitivity analyses , excluding patients without splenomegaly on admission , to assess the degree of mis-diagnosis . ( 2 ) The transition from SSG to SSG/PM in the second half of the period in our study was accompanied by an increase in the proportion of VL relapse patients for whom a record of previous treatment could be identified . In order to assess possible bias in estimating the association of SSG/PM vs SSG with risk of relapse , we performed a sensitivity analysis in which we assumed that VL relapse patients for whom a record of previous treatment could not be identified ( hence , who were excluded from the main analysis of risk factors for relapse ) had the same characteristics on admission for primary VL ( same age , sex , treatment centre , time of treatment and treatment with SSG or SSG/PM ) as they did on admission for relapse VL . Data were collected as part of routine patient care; no additional investigations were performed . Ethical approval was obtained from the MSF Ethical Review Board .
The treatments for which electronic archive data were available are summarised in Table 1 . Overall , records were available for 78 . 2% ( 10 , 105/12 , 924 ) of treatments , according to MSF monthly field activity reports . For relapse VL , the corresponding proportion of treatments with entered data was higher , at 88 . 1% ( 621/705 ) ( Figure 1 ) . The Lankien and Pieri treatment centres were looted in 2004 and 2006 , respectively , with the loss or destruction of many treatment cards . Other losses of treatment cards were due to generally adverse conditions in the field and difficulties in collecting and transporting paper records from remote sites , some of which had no permanent expatriate presence . Our analysis was based on primary VL patients who were known to have subsequently relapsed because a record of treatment for VL relapse could be linked with their primary VL treatment record ( Figure 1 ) . These relapse patients ( N = 166 ) were a subset ( 26 . 7% ) of all VL relapse patients who had an electronic record of treatment for relapse VL ( N = 621 ) . Table 2 shows a comparison of the characteristics of VL relapse patients who did or did not have a record of previous treatment for primary VL . Patients for whom we could not identify a previous treatment record had more missing data , higher adult BMI , longer duration of illness , and less splenomegaly than patients whose previous treatment data were used in our analysis of factors associated with VL relapse , but there were no differences in age , sex , nutritional status of patients ≤19 years old , haemoglobin , walking status , drug treatment , or treatment outcome . More than half ( 57 . 8% ) of the 621 VL relapse patients were treated at one centre ( Lankien ) , and a record of previous treatment was available for 34 . 3% ( 123/359 ) of these patients compared with 16 . 4% ( 43/262 ) for all other treatment centres combined ( P<0 . 001 ) . A record of previous treatment was more likely to be identified in the second half of the period under study ( 36 . 5% in 2003–2007 ) compared with the first half ( 14 . 1% in 1999–2002 ) ( P<0 . 001 ) . Table 3 shows a comparison of the characteristics on admission of primary VL patients who did or did not have a recorded subsequent relapse . Higher odds of relapse were associated with splenomegaly ( on admission for , and discharge from , treatment for primary VL ) and with SSG/PM treatment of primary VL . In a logistic regression model which included age , sex , year and treatment centre , admission spleen size ( Hackett grade ) ≥3 was associated with >4-fold higher odds of relapse compared with size 0 ( odds ratio ( OR ) = 4 . 40 ( 95% CI 1 . 74–11 . 08 ) , P = 0 . 002 ) . Admission spleen grades 1 and 2 were associated with >2-fold and 3-fold higher odds: OR = 2 . 43 ( 0 . 95–6 . 26 ) , P = 0 . 02 for size 1 vs size 0; OR = 2 . 91 ( 1 . 17–7 . 24 ) , P = 0 . 07 for size 2 vs size 0 . In a similarly-adjusted model , 17-day SSG/PM was associated with >2-fold higher odds of relapse ( OR = 2 . 26 ( 1 . 46–3 . 51 ) , P<0 . 001 ) compared with 30-day SSG monotherapy . This estimate was robust to a sensitivity analysis in which the N = 455 VL relapse patients for whom no record of previous treatment was available were included in the dataset as primary VL patients who subsequently relapsed , under the assumption that these patients were treated with the same drug for both VL episodes ( OR = 2 . 25 ( 1 . 79–2 . 82 ) , P<0 . 001 ) . Young age ( <5 years ) was not associated with risk of relapse . However , among children <5 years old , infancy ( age <1 year ) compared with age 1–4 years was associated with higher risk of relapse ( OR = 3 . 47 ( 1 . 16–10 . 37 ) , P = 0 . 03 ) . Patients treated at Lankien were more than twice as likely to relapse ( OR = 2 . 32 ( 1 . 61–3 . 36 ) , P<0 . 001 ) as patients treated at any other centre ( model adjusted for age , sex and year of treatment ) . Test-of-cure data were available for only 4 . 2% ( 7/166 ) of relapse patients who had a previous treatment record vs 5 . 4% ( 431/7 , 924 ) of patients with no record of subsequent relapse ( P = 0 . 6 ) . None of the final test-of-cure results for relapse patients , and only 3/431 of the test-of-cure results for patients with no record of subsequent relapse were positive . Hence , test-of-cure data were not included in the risk factor analysis . Whether or not a test-of-cure was performed ( indicative of non-response to treatment ) was not associated with risk of relapse ( OR = 1 . 19 , 95% CI 0 . 60–2 . 34 ) . Patients who did not have a recorded relapse were missing more height and weight data , but none of the anthropometric measures on admission were associated with VL relapse . Although mean hemoglobin was not associated with VL relapse , in multivariable regression ( adjusted for age , sex , year and treatment centre ) primary VL patients presenting with hemoglobin in the lowest quartile ( ≤7 . 0g/dl ) had an odds ratio for relapse of 2 . 60 ( 1 . 00–6 . 74 ) , P = 0 . 05 ) compared with patients in the highest quartile ( >10g/dl ) . Splenomegaly and anemia were strongly correlated ( decrease in Hb concentration for each increment in Hackett grade = 0 . 47 ( 0 . 40–0 . 54 ) g/dl , P<0 . 001 , by linear regression adjusted for age and sex ) . Higher risk of relapse was also strongly associated with spleen size at end of treatment ( OR = 5 . 41 ( 1 . 89–15 . 47 ) , P = 0 . 002 for size ≥3 vs size 0; OR = 2 . 09 ( 1 . 19–3 . 67 ) , P = 0 . 01 for size 2 vs size 0; OR = 1 . 45 ( 0 . 82–2 . 53 ) , P = 0 . 2 for size 1 vs size 0 ) ( Table 4 ) . These associations were slightly attenuated if we excluded patients who had no palpable spleen or no spleen size data on admission ( OR = 5 . 01 ( 1 . 75–14 . 37 ) , P = 0 . 003 for size ≥3 vs size 0; OR = 1 . 98 ( 1 . 12–3 . 48 ) , P = 0 . 02 for size 2 vs size 0; OR = 1 . 31 ( 0 . 74–2 . 34 ) , P = 0 . 4 for size 1 vs size 0 ) . Spleen size at end of treatment for primary VL was not correlated with spleen size on admission for relapse VL ( pairwise correlation coefficient = 0 . 12 , P = 0 . 2 ) . There were gains in BMI and WFH during treatment among patients who did and who did not have a subsequently recorded relapse , although these gains were statistically evident ( P<0 . 001 ) only for the ( much larger ) group of non-relapsing patients . Weight on discharge was recorded for only 50% of patients . Among patients of all ages , the mean gain in body weight ( as a percentage of weight on admission ) was higher in non-relapsing patients ( 4 . 2% ) than in relapsing patients ( 2 . 4% ) . In a linear regression model ( adjusted for age , sex , year , treatment centre , and weight on admission ) , the difference in percentage weight gain between non-relapsing and relapsing patients was 2 . 29% ( 95% CI 0 . 43%–4 . 14%; P = 0 . 02 ) . In a multivariable model ( Table 5 ) comprising age , sex , year , treatment centre , spleen size on admission , spleen size on discharge , and treatment given ( SSG/PM vs SSG ) : splenomegaly Hackett grade ≥3 on admission for , and on discharge from , treatment for primary VL were independently associated with higher odds of subsequent relapse compared with size 0 ( on admission OR = 3 . 62 ( 1 . 08–12 . 12 ) , P = 0 . 04; on discharge OR = 5 . 50 ( 1 . 84–16 . 49 ) , P = 0 . 002 ) ; SSG/PM was associated with higher odds of relapse ( OR = 2 . 08 ( 1 . 21–3 . 58 ) , P = 0 . 008 ) ; and patients treated at Lankien were more likely to relapse ( OR = 1 . 76 ( 1 . 14–2 . 72 ) , P = 0 . 01 ) than patients treated at any other centre . In a sensitivity analysis , excluding patients who had spleen size 0 or missing spleen size data on admission from this model ( Table 5 ) , the association of splenomegaly on discharge with higher odds of relapse was slightly attenuated ( OR = 5 . 38 ( 1 . 79–16 . 12 ) , P = 0 . 003 for size ≥3 vs size 0 ) , as was the association of SSG/PM with higher odds of relapse ( OR = 2 . 02 ( 1 . 16–3 . 51 ) , P = 0 . 01 ) . SSG/PM compared with SSG monotherapy was associated with 73% lower odds of death among primary VL patients ( OR = 0 . 27 ( 0 . 20–0 . 37 ) , P<0 . 001 ) and 85% lower odds of death among relapse VL patients ( OR = 0 . 15 ( 0 . 05–0 . 46 ) , P = 0 . 001 ) , in models adjusted for age , sex , year and treatment centre . Table 6 shows summary statistics for the proportion of patients admitted to MSF treatment centres in Southern Sudan who were diagnosed with VL relapse ( expressed as a proportion of patients treated for primary VL ) . Join-point analysis provided weak evidence ( P = 0 . 12 ) for an upward trend in this proportion ( annual percent change ( APC ) = 11 . 4% ( −3 . 4% to 28 . 5% ) ) , as illustrated in Figure 2 . The proportion of patients who died fell dramatically between 1999 and 2001 ( from 13 . 5% to 7 . 4% ) and then declined steadily to <3% in 2006/2007 . Join-point analysis provided strong evidence ( P<0 . 001 ) for a downward trend in deaths during treatment for VL over the period 1999–2007 ( APC = −18 . 1% ( −22 . 5% to −13 . 4% ) , P<0 . 001 ) ( Table 6 , Figure 2 ) . These trends in relapse and deaths were corroborated by logistic regression models ( including age , sex , and treatment centre ) , which showed no discernible trend in risk of relapse ( OR = 1 . 04 ( 0 . 95–1 . 14 ) , P = 0 . 4 ) , but a clear decline in risk of death ( OR = 0 . 81 ( 0 . 77–0 . 86 ) per year , P<0 . 001 ) . The proportion of primary VL patients presenting with splenomegaly indicated by Hackett Grade ≥3 decreased from 34 . 4% in 1999 to 16 . 0% in 2007 ( Table 6 , Figure 2 ) ; a trend of −9 . 6% ( −13 . 5% to −5 . 5% ) per annum ( P<0 . 001 ) . The mean time-to-presentation for primary VL patients fell from 2 . 8 to 1 . 2 months , and for relapse patients from 3 . 0 months in 1999 to 1 . 6 months in 2007 ( Table 6 ) . Join-point analysis showed a steady downward trend in the mean time-to-presentation of VL relapse patients ( APC = −6 . 7% ( −11 . 3% to −1 . 8% ) , P = 0 . 003 ) and of primary VL patients ( APC = −7 . 7% ( −3 . 8% to −11 . 6% ) , P = 0 . 003 ) ( Table 6 , Figure 2 ) . Figure 2 also shows that the proportion of primary VL patients treated with SSG/PM combination therapy , instead of SSG monotherapy , increased from <1% in 2001 to 90% in 2003 ( Table 6 ) . We found no trends in age and sex or in nutritional and walking status . The median interval between discharge from treatment for primary VL and re-admission for VL relapse was 91 days ( range 19–1358 days , N = 166 ) ( Table 6 ) ; the median interval in years 1999–2003 ( 105 days ) was longer ( by 27 days ) than in years 2004–2007 ( Kruskal-Wallis test , 1df , P = 0 . 04 ) . The overall proportion of patients relapsing within 3 months was 49 . 4%; within 6 months 83 . 7%; and within 12 months 96 . 4% . The interval between discharge from treatment for primary VL and re-admission for relapse was not associated with age , sex , spleen size on discharge , nutritional status on discharge or drug regimen .
We have shown that splenomegaly among patients with primary L . donovani VL was strongly associated with a much higher relative risk of VL relapse . Primary VL patients presenting with splenomegaly of Hackett Grade ≥3 had almost four-fold higher odds of subsequent relapse than patients with no enlargement of the spleen , and patients discharged with greatly enlarged spleens had greater than five-fold higher odds of relapse . The latter , but not the former , association remained evident after excluding from our analysis all patients who had no palpable spleen on admission ( Table 5 ) . No other clinical characteristics of primary VL patients emerged as strong risk factors for VL relapse , although infants appeared to be more susceptible as reported for L . chagasi VL [19] . Splenomegaly on admission may indicate a combination of severity of illness , parasite burden , and severity of immunosuppression . Splenomegaly on discharge suggests the patient has not responded adequately to treatment , and may either harbour a significant parasite burden or may still be immunosuppressed by the disease . Splenomegaly is one of the classical signs of VL , reported globally as present in >90% of VL patients [13] . Extremely enlarged spleen on admission is a sign of advanced disease and is a risk factor for death [3] . The spleen in L . donovani infection is infiltrated by parasitized macrophages as well as plasma cells , immune complexes and other components of immune response , leading to hyperplasia of reticulo-endothelial cells and enlargement of the organ [20] . A chronic inflammatory state mediated mainly by TNF results in architectural damage and immunological dysfunction [21] , [22] . Particularly important for treatment outcome is loss of spleen marginal zone macrophages , which play an important role in capturing blood-borne pathogens; during enlargement , the spleen's protective role is reduced [23] , [24] . Through clinical and pathophysiological observation , reduction of spleen size is recognised as one of the most important signs of successful treatment , and splenectomy has traditionally been practiced in splenomegalic patients who have frequent relapses despite appropriate treatment [25] . However , according to the WHO Manual on Visceral Leishmaniasis , a completely unpalpable spleen is not considered necessary to classify a patient as cured , ( “persistently enlarged spleen is no cause for concern provided the patient's other symptoms are improving” ) [13] . Severe anaemia is associated with the increased risk of death among VL patients [3] and , apart from other reasons ( bone marrow suppression , malnutrition and generalised inflammation ) , is related to intensive destruction of erythrocytes in enlarged spleens [24] . Thus , the correlation between marked splenomegaly and anaemia found in our data would be expected . However , other indicators of advanced disease on admission , such as severe malnutrition and general weakness expressed as walking status , were not associated with relapse in our patients . Weight gain is a sign of clinical response to treatment , and mean percentage weight gain was marginally lower in the group of patients who experienced relapse . However , we note that discharge weight data were missing for 50% of patients . The mean percentage gain in haemoglobin level was the same in both groups , although data were missing for 80% of patients . Other signs of clinical response , such as improvement of general status and becoming afebrile , were not recorded in our database . We found that 17-day SSG/PM combination therapy was associated with two-fold higher odds of relapse than 30-day SSG monotherapy . The joinpoint analysis showed an upward trend in VL relapse admissions as a proportion of primary VL admissions ( Figure 2 ) , which might be interpreted as a consequence of the introduction of SSG/PM as first-line treatment for primary VL between 2001 and 2003 . However , this apparent trend was driven by the relatively high proportion of VL relapse patients treated in the last two years ( 2006–2007 ) , and the risk factor analyses showed no discernible trend . The 17-day SSG/PM regimen was introduced on the basis of evidence from several trials [26]–[28] , and in a retrospective study ( 2002–2005 ) based on MSF treatment data from Southern Sudan , the SSG/PM regimen was found to give better survival and cure rates than SSG alone [9] . Preliminary analysis of a phase III clinical trial by the Drugs for Neglected Diseases Initiative ( DNDi ) indicated as yet no difference in cure rates between 17 days of SSG/PM and 30 days of SSG [DNDi - unpublished data]; final results of the trial are due in 2010 [29] . It could be that the very patients who were saved from dying by SSG/PM may be those who relapsed . This phenomenon probably occurred during our study of miltefosine vs SSG in Ethiopia: miltefosine , being a safer drug than SSG , was associated with a far lower death rate , yet a far higher relapse rate [30] . However , not all previously-reported risk factors for death were evident as risk factors for relapse ( possibly because drug toxicity is a primary cause of death ) [3] , and we cannot discount a possible role of shorter duration of treatment in increasing risk of relapse [31] . Our reported association between SSG/PM and risk of VL relapse was likely to be residually confounded , and possibly biased by missing data , because SSG/PM was used at permanent MSF treatment centres , whereas SSG monotherapy was used at temporary ( seasonal ) outreach sites . Access to care in the event of VL relapse was much easier at the permanent sites , some of which have even benefited from public transport since 2005 . Temporary sites were also less able to diagnose relapses , because the splenic or lymph node aspirates which are necessary to make the diagnosis of relapse could only be carried out at the main treatment centres . These factors , together with better record-keeping , probably explain the apparently higher risk of relapse associated with treatment at Lankien . The use of SSG/PM during the latter half of the period in our study coincided with a series of political agreements ( between 2003 and 2005 ) , which brought a tentative ceasefire to Southern Sudan after decades of conflict . Hence , the crude upward trend seen in MSF's summary data is probably a consequence of easier access to treatment , as reflected in the trend towards shorter ‘time-to-presentation’ for VL patients ( Figure 2 ) . In the past , when access to care was restricted , patients with relapse may have died without reaching a treatment centre . HIV infection was unlikely to be a factor in VL relapse during most of the period in our study . An unlinked screening study of 206 VL patients in Lankien in 2002 revealed only 1 HIV-positive case ( 0 . 5% ) ; or 1 in 36 adult patients ( 2 . 8% ) [MSF - unpublished data] . However , the number of VL patients has declined in recent years ( from 4 , 172 in 2003 to 101 in 2008 ) , while large numbers of refugees have begun to return to Southern Sudan from places with higher HIV prevalence ( Ethiopia , Kenya , Khartoum ) . Hence we might expect the proportion of VL relapses attributable to HIV co-infection to increase . Recent ( 2008 ) data from Nasir , where MSF began routine HIV testing and anti-retroviral therapy in adult VL patients , showed that 25% ( 5/20 ) were HIV co-infected [MSF - unpublished data] . Our study was the largest retrospective analysis of VL relapse globally . The difficult conditions under which the data were collected gave rise to some limitations , principally missing treatment records , missing data and , for this study , our inability to link all VL relapse patients with their record of treatment for primary VL . This was mainly because the majority of patients did not present the identification card given to them when they were discharged from treatment for primary VL . It is difficult to assess the impact of missing records and missing data on our outcomes , although we have no reason to suspect these were a source of significant bias in our analysis of clinical risk factors . Some of the primary VL patients whom we classified as “primary VL patients who did not subsequently relapse” may have had an untreated or unrecorded relapse , and some patients who had been previously treated may have been diagnosed as primary VL if the patient did not recall their primary episode . These two misclassifications would lead to under-estimation of risk factors for relapse . The 5 patients ( 3% ) re-admitted soon after discharge ( 19 to 31 days ) may have been treatment failures , rather than relapses . Through our sensitivity analysis we attempted to adjust for those patients with no palpable spleen , who were possibly falsely diagnosed as VL patients ( hence at zero risk of relapse ) . However , in a setting with high prevalence of other endemic diseases which cause splenomegaly ( malaria , typhoid , schistosomiasis , brucellosis , and liver cirrhosis ) , some patients with splenomegaly may also not have had primary VL . We could not conduct a further sensitivity analysis , based on the actual diagnostic method for each patient ( parasitological , immunological , clinical ) , because this could not be deduced from the available data . We noted that relapse VL patients for whom we could not identify a previous treatment record had less severe splenomegaly on re-admission than relapse VL patients whose previous treatment data were used in our analysis . However , spleen size on admission for relapse VL was not correlated with spleen size at end of treatment for primary VL , hence this difference would not have led to an over-estimate of the association between splenomegaly at end of treatment for primary VL and subsequent relapse . Our finding that SSG/PM is associated with increased risk of relapse is likely to be confounded by improvements in access to treatment which coincided with the introduction of this shorter combination drug regimen , and may reflect better survival rates compared with SSG monotherapy . We await a definitive answer from a randomised controlled trial of SSG/PM vs SSG currently underway in East Africa . Meanwhile , our finding that splenomegaly is associated with increased risk of VL relapse could contribute to revised guidelines for clinical assessment of VL patients prior to discharge .
|
Visceral leishmaniasis ( kala-azar ) caused by Leishmania donovani is spread from person to person by Phlebotomus sandflies . Major epidemics of visceral leishmaniasis have occurred in Southern Sudan during the 20th century . The worst of these killed 100 , 000 people in the western Upper Nile area of Southern Sudan from 1984–1994 , a loss of one-third of the population . Médecins Sans Frontières has treated 40 , 000 kala-azar patients in Southern Sudan since the late 1980's . In this study we used routinely collected clinical data to investigate why some patients relapse after treatment . We found that patients with severely enlarged spleens ( splenomegaly ) are more likely to relapse . Patients treated for 17 days with a combination of two drugs ( sodium stibogluconate and paromomycin ) were more likely to relapse ( but less likely to die ) than patients treated for 30 days with a single drug ( sodium stibogluconate ) . However , the transition from sodium stibogluconate to the sodium stibogluconate/paromomycin combination as standard treatment between 2001–2003 has not led to a significant increase in visceral leishmaniasis relapse .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"public",
"health",
"and",
"epidemiology/epidemiology",
"public",
"health",
"and",
"epidemiology/infectious",
"diseases",
"infectious",
"diseases/protozoal",
"infections",
"infectious",
"diseases/epidemiology",
"and",
"control",
"of",
"infectious",
"diseases"
] |
2010
|
Visceral Leishmaniasis Relapse in Southern Sudan (1999–2007): A Retrospective Study of Risk Factors and Trends
|
The Ori region of bacterial genomes is segregated early in the replication cycle of bacterial chromosomes . Consequently , Ori region positioning plays a pivotal role in chromosome dynamics . The Ori region of the E . coli chromosome is organized as a macrodomain with specific properties concerning DNA mobility , segregation of loci and long distance DNA interactions . Here , by using strains with chromosome rearrangements and DNA mobility as a read-out , we have identified the MaoP/maoS system responsible for constraining DNA mobility in the Ori region and limiting long distance DNA interactions with other regions of the chromosome . MaoP belongs to a group of proteins conserved in the Enterobacteria that coevolved with Dam methylase including SeqA , MukBEF and MatP that are all involved in the control of chromosome conformation and segregation . Analysis of DNA rings excised from the chromosome demonstrated that the single maoS site is required in cis on the chromosome to exert its effect while MaoP can act both in cis and in trans . The position of markers in the Ori region was affected by inactivating maoP . However , the MaoP/maoS system was not sufficient for positioning the Ori region at the ¼–¾ regions of the cell . We also demonstrate that the replication and the resulting expansion of bulk DNA are localized centrally in the cell . Implications of these results for chromosome positioning and segregation in E . coli are discussed .
The size of genomes with respect to cellular dimensions imposes the need for extensive chromosome condensation that is compatible with the genome replication and the expression of genetic information . The bacterial chromosome is organized at multiple levels from the genome content ( including the distribution of DNA motifs , genes and replication arms ) , to chromatin composition , supercoiling domains and large chromosomal domains [1 , 2] . The E . coli chromosome is divided into four large domains called macrodomains ( MDs ) [3 , 4] . MDs are defined as large regions in which DNA interactions occurred preferentially , whilst DNA interactions between different MDs are restricted . The Ori MD contains the origin of replication oriC , the opposite Ter MD contains the replication terminus and the chromosome dimer resolution dif site , whilst the Left and Right MDs flank the Ter MD . Two non-structured ( NS ) regions ( NSRight and NSLeft ) flank the Ori MD . DNA sites within the NS regions can interact with both flanking MDs [4] . MD organization influences the segregation of sister chromatids and the mobility of chromosomal DNA [5] . Each MD occupies a specific territory inside the nucleoid and is segregated with specific properties [5] . The molecular details of Ter MD organisation and segregation have previously been uncovered . The protein MatP ( Macrodomain Ter protein ) binds to a specific DNA sequence , termed matS , that is repeated 23 times within the 800 kb Ter region [6 , 7] . Furthermore , the interaction of MatP with the division apparatus associated protein ZapB promotes the anchoring of the Ter MD at mid-cell to control segregation of the Ter MD [8] . The mobility of DNA markers in MDs is highly reduced compared to NS regions . In the Ter MD , constraints on DNA mobility resulted both from the binding of MatP to matS sites and from its interaction with the divisome . The mechanisms responsible for structuring and constraining DNA mobility in other MDs have not been described . The understanding of bacterial chromosome segregation has improved considerably in recent years [1] . The Ori region from bacterial chromosomes plays a pivotal role in chromosome organization and segregation as it is replicated and segregated early in cell division and its positioning impacts the cellular organization of the chromosome in the cell [2] . In many bacterial species , ParABS play a critical role in the condensation and segregation of the Ori region . Although present on different low copy number plasmids in E . coli ( e . g . F , P1 ) [9] , ParABS or any analogous system has not been found in the chromosome of E . coli or other enteric bacteria . In E . coli , a specific cis-acting site called migS affects bipolar positioning of oriC but its deletion does not cause any severe defects in chromosome partitioning [10] . Different processes have been proposed to contribute to the partitioning and segregation of the E . coli chromosome , including the radial confinement of nucleoid DNA [11] , migS-driven separation of Ori regions [10] , entropic exclusion of sister chromosomes [12] , differential condensation levels of the chromosomal regions [13] and a MukB-TopoIV interaction for decatenation and segregation of newly replicated Ori DNA [14] . The final destination of the Ori region is controlled and maybe responsible for the transversal organization of the chromosome in these species [15–17] . However , the mechanisms that control the positioning of the Ori region remain undefined . Genomic rearrangements have previously been used to unravel the principles of chromosome organization [18–25] . In many cases , the recombination between inverted recombining sites was used to promote the rearrangement either using general recombination systems or site specific recombination modules . Although this method can be applied with sites dispersed all over the genome it can unwittingly affect several parameters at the same time . For example , inversion of a segment encompassing the Right and Ori MD will not only generate hybrid MDs but will also affect the orientation of genes and sequences inside hybrid MDs and the intervening segment generating possible replication-transcription conflicts [26] . To limit potential replication-transcription conflicts resulting from genetic rearrangements , we previously developed a strategy that allows the transposition of any chromosomal segment at any position in the chromosome in any orientation . Thus one can change the position of any DNA segment whilst preserving its orientation relative to the replication process [27] . These rearrangements rely on a cut-and-paste process reminiscent of a transposition event and involve the site-specific recombination module of bacteriophage λ . In an attempt to define the mechanism responsible for structuring the Ori MD , this current study has generated new chromosome configurations by transposing different regions of the Ori MD , Right MD or NSRight region without affecting gene orientation . By analyzing the mobility of transposed segments we show that different regions of the Ori MD obtained new mobility properties when present in a different genomic context . Interestingly , a specific 17 bp sequence from the Ori region , termed maoS , was responsible for restricting DNA mobility in adjacent regions . A gene adjacent to maoS , termed maoP ( previously known as yifE ) was also required for constraining marker mobility in the Ori region . Inactivation of maoP removed constraints on DNA mobility in the Ori MD and allowed long-range interaction between Ori and Right MDs . The maoS/MaoP system contributed to the control of Ori region choreography , as inactivation of maoP affected the position of Ori markers . We also demonstrate that DNA replication and the resulting expansion of bulk DNA positions the nucleoid in the middle of the cell . Implications of these results for nucleoid positioning and chromosome conformation are discussed .
Our first goal was to understand if sequences within the Ori domain influence DNA mobility . To do this , we monitored movement of several chromosomal regions . The mobility of foci was determined by measuring the distance traveled by a focus from the home position at 10 sec intervals for a period of 5 min [5] . The motion of these regions was compared before and after chromosomal segments had been displaced by transposition . The transposed ( Td ) segments are labeled according to their original position with a numerical suffix , i . e . segments from the Ori region are called “ori Td-1” , “ori Td-2” and etcetera . A line below the genetic map indicates the transposed segments , and an arrowhead represents the insertion site . Different segments from the Ori and Right MDs or from the NSRight region were inserted in the Right or Ori regions ( Fig 1 and S1 Fig ) . When two segments proximal to oriC ( i . e the 138-kb ori Td-1 ( Fig 1A ) and the 302-kb ori Td-2 ( S1A Fig ) ) were transposed to the Right MD at 17’ , markers in the transposed Ori segment ( Ori-6 and Ori-7 ) showed the same constraints on DNA mobility as in the original configuration ( Fig 1A and S1A Fig ) . In this configuration , markers in the Right MD ( Right-5 ) and in the NSRight region ( NSR-5 ) also showed the same mobility as in the WT context . In contrast , other markers in the Ori region not involved in the transposition event ( markers Ori-3 and Ori-4 ) displayed a higher mobility ( Fig 1A ) . Transposition of segments distal to oriC ( i . e the 318kb ori Td-4 ( Fig 1B ) and the 143kb ori Td-3 ( S1 Fig ) ) to the same locus in the Right MD resulted in transposed Ori markers ( Ori-3 and Ori-4 ) with an increased mobility at their new location ( Fig 1B and S1B Fig ) . The mobility of Ori markers not involved in the transposition remained unchanged . Interestingly , the marker NSR-1 in the NSRight region , that was positioned closer to oriC , had a reduced mobility compared to the WT context . Altogether , these data indicate that the Ori region is composed of heterogeneous parts that have altered properties in different genetic contexts . In conclusion determinants responsible for limiting DNA mobility are not scattered homogeneously through the Ori region . Instead they are present in the ori Td-1 segment and constraints on DNA mobility are spread to adjacent segments . The results described above suggested that constraints in the Ori region can spread to NS markers if they are brought closer to C . To test this hypothesis , a 264kb segment from the NSRight region ( NSR Td-2 ) was transposed to two different positions ( respectively 92 . 7 min and 87 . 7 min ) in the Ori region ( Fig 1C and S1C Fig , respectively ) . At both positions , the mobility of markers NSR-1 and NSR-2 in the transposed segment was constraint whilst the mobility of a marker at its original position ( NSR-5 ) was not affected ( Fig 1C and S1C Fig ) . The mobility of markers originally located in the Ori region ( Ori-3 and Ori-4 ) were increased following transposition even though they did not acquire the mobility of a NSR marker located the same distance from oriC in the WT configuration ( Fig 1C ) . Finally , analysis of NSR-2 and NSR-5 mobility in a configuration where a large segment from the NSRight region ( NSR Td-1 ) is transposed within the Right MD ( Fig 1B ) indicated that all markers displayed the same mobility as in the WT strain . In conclusion the insertion of segments from either the Ori region or the NSRight region into the Right MD had no consequences on the mobility of loci of the Right MD . To determine the consequences of transposing a segment of the Right MD into the Ori region , we analyzed the mobility of various markers originally located in the Right and Ori MDs following chromosome rearrangement . A 154-kb segment ( right Td-1 ) ( Fig 1D ) and a 293-kb segment ( right Td-2 ) ( S1D Fig ) were transposed into two different positions ( respectively 92 . 7 min and 87 . 7 min ) in the Ori MD . The mobility of markers in the transposed segment from the Right MD was unchanged . Interestingly the Ori markers ( Ori-3 and Ori-4 ) that were separated from oriC by the transposed Right MD segment Right Td-1 had increased mobility ( Fig 1D ) . Locus Ori-1 , whose position remained unchanged , had similar constraints on DNA mobility as in the WT context ( Fig 1D ) . Similar results were obtained for markers Right-5 , Ori-3 and Ori-4 after insertion of Right Td-2 at 92 . 7 and 87 . 6 min ( S1D Fig ) . In conclusion , segments from the Right MD were able to impede the transmission of mobility constraints originating from the region proximal to oriC . By utilizing the ability of transposed Right MD segments to increase the mobility of Ori MD markers , we attempted to map the determinants required for constraining DNA in the Ori MD . By inserting a segment from the Right MD at different positions in the Ori region and measuring the mobility of markers on either side of the transposed segment , it was predicted that marker mobility would be affected relative to their position to DNA constraining determinants . The 293-kb right Td-2 segment was transposed at five positions in the Ori region and marker mobility was assessed at the five positions: 87 , 7 min ( a ) , 86 . 7 min ( b ) , 86 . 1 min ( c ) , 85 . 1 min ( d ) and 84 . 7 min ( e ) ( Fig 2A ) . Insertion of right Td-2 at positions a , b , c , and d resulted in an increased mobility of markers distal from oriC ( i . e . Ori-3 and Ori-4 ) whereas a marker close to oriC ( i . e . Ori-5 ) was not affected ( Fig 2A ) . When right Td-2 was inserted at 84 . 7 min ( e position ) the mobility of markers distal to oriC was not affected whilst the marker close to oriC had a higher mobility ( Fig 2A ) . In conclusion , determinants responsible for constraining the Ori region were present in a region of 69-kb ( coordinates 3928200 to 3998000 ) between positions d and e . To precisely map the position of determinants constraining DNA mobility , four deletions corresponding to non-overlapping segments varying in size between 2 kb and 17 kb were generated in the 69 kb region ( Fig 2B ) . Two regions could not be deleted because of the presence of essential genes . The deletion of the 2 kb segment ( delta 15–17; coordinates 3944500–3946500 ) resulted in an increased mobility of the Ori MD markers ( Ori-3 and Ori-4 ) , whilst the three remaining deletions had no substantial effect ( Fig 2B ) . This region contains three genes: hdfR , yifE and yifB . hdfR and yifE are divergent genes separated by a 118-bp long intergenic region and yifB is convergent with yifE ( Fig 2C ) . To further define the determinants required for conferring Ori MD properties , additional deletions were performed and the mobility of markers Ori-3 , Ori-4 and NSR-2 were measured ( Fig 2C ) . These results demonstrate that both the yifE gene and the upstream intergenic region ( IR ) were required to constrain the mobility of markers Ori-3 and Ori-4 . It is interesting to note that deletion of hdfR had an opposite effect as the mobility of all three markers was more constrained in this context . The deletion of yifB had no effect on marker mobility . In conclusion , yifE and the upstream intergenic region ( IR ) are required for constraining DNA mobility in the Ori MD . To characterize the determinants required for limiting mobility of Ori markers , constructs were generated to provide the IR and yifE in trans ( i . e . present on a self-replicating pSC101 plasmid derivative ) in different chromosomal contexts , i . e . strains deleted for yifE and the IR , or carrying the IR at different loci . The IR region presumably contains the promoter for yifE . The presence of yifE and the upstream IR on a replicating plasmid was not able to constrain the mobility of markers in the region ( Ori-3 and Ori-4 ) when yifE and IR were deleted from the chromosome ( Fig 3A ) . In contrast , the presence of the IR on the chromosome , either at the WT position or at NSR-1 position , together with the replicating plasmid carrying yifE and IR restored the constrained mobility of Ori-3 and Ori-4 markers in a yifE deletion mutant ( Fig 3A ) . Interestingly , the mobility of the NSR-2 marker was constrained when the IR was inserted in the NS region ( NSR-1 ) in the presence of the plasmid carrying yifE . These results demonstrate that the IR exerts its constraint on DNA mobility both in its left and right , and must be present in cis , whilst yifE can act in trans or in cis , either when present on the plasmid or inserted in the chromosome . To identify the determinants required in cis , various part of the IR were tested for their ability to constrain DNA mobility of marker NSR-2 when inserted at position NSR-1 . We first identified a 30 nts segment in the middle of the IR and subsequently , a 17 nts long motif with the sequence CTAATACTCCGCGCCAT was shown to limit the mobility of marker NSR-2 ( S2 Fig ) . This motif was termed maoS ( Macrodomain Ori Sequence ) . To confirm that maoS may act in cis to constrain mobility of DNA markers , we used a site-specific recombination system to loop out various parts of the Ori region from the chromosome and analyzed the subsequent effects on marker mobility . The system relies on the λ Int and Xis recombination genes . When the site-specific recombination attR and attL sites are present in a direct orientation , the recombinase promotes the excision of the intervening segment resulting in the formation of cells carrying a chromosome split in two parts ( Fig 3B ) . As only one of the parts carries oriC , the cells will not be viable if essential genes are present on the non-replicated ring . The efficiency of excision varies relative to the distance separating the two att sites and can be assessed by plating for viable cells as only cells that did not sustain an excision event will be viable . Reinsertion of the excised ring does not occur at a high frequency [27] . To visualize the fate of excised rings immediately following excision and to measure the mobility of markers , a parS tag targeted by ParB-GFP was inserted on the excisable segment . The segments ori Td-1 and ori Td-3 were excised from the chromosome and the mobility of markers either present on the chromosome or on the non-replicated ring were assessed ( Fig 3C and 3D ) . The markers located in cis with maoS ( Ori-3 , Ori-4 , Ori-6 in the WT context , Ori-6 upon excision of ori Td-1 and Ori-4 in the context of ori Td-3 ) were highly constrained ( Fig 3C and 3D ) . In contrast , the mobility of markers not in cis with maoS ( Ori-3 upon ori Td-1 excision and Ori-3 upon ori Td-3 excision ) was increased following excision . Combined , these results indicate that constraints in the Ori region resulted from the presence of maoS exerting its effect in cis over large distances either in the chromosome or in excised rings . In conclusion , these results identified two elements required for constraining DNA in the Ori MD: the hdfR-yifE IR acting in cis and likely acting as a target sequence and the yifE gene acting in trans likely through the encoded product . The yifE gene was named maoP ( Macrodomain Ori Protein ) . Inactivation of MaoP/maoS generated a number of abnormal cells ( including ~5% of elongated cells ) with a defect in nucleoid segregation ( S3 Fig ) . In minimal medium supplemented with casamino acids or in LB medium at 30°C , the growth rate is slightly affected in a maoP mutant and the generation time is 100 min and 55 min , respectively ( compared to 85 min and 50 min in WT cells ) . Control of replication initiation is also affected upon MaoP inactivation ( S3 Fig ) ; in minimal media , replication initiation is delayed in maoP cells compared to wt cells ( about 50% of cells with 4 copies of oriC in maoP cells compared to more than 2/3 of cells with 4 copies of oriC in wt cells ) . The Ori region has two properties that define it as a MD [4 , 5] . First , the mobility of Ori markers is lower than those of flanking NS regions and second , interactions of loci present in the Ori and Right MDs occur at a low frequency . Large intrareplichore inversions , that intermingled Ori and Right MDs , occur rarely and results in a configuration that is detrimental for growth [25] . These properties of the Ori MD were analyzed upon inactivation of the MaoP/maoS system . Firstly , the mobility of markers located in the Ori region were affected whilst markers in the NS region , the Right MD or the Ter MD exhibited a mobility similar to that observed in WT cells ( Fig 4A ) . Secondly , DNA inversions between Ori and Right MDs were not detected or occurred at a very low frequency in WT cells ( Fig 4B and Table 1 ) . However , the percentage of recombinants in Ori-Right combinations was increased in maoP deletion mutants ( Fig 4B and Table 1 ) . This increase is specific for interactions between the Ori and Right MDs as no differences were seen for interactions between the Right and Left MDs or between the Right and Ter MDs ( Table 1 ) . Furthermore , maoP inactivation increased the size of the colonies with the inverted chromosomes with respect to inversions in a WT background ( Fig 4B ) . Combined , the results demonstrate that MaoP specifies Ori MD properties including the constraint mobility of Ori markers and restrictions on long distance interactions with loci in the Right MD . To determine if MaoP/maoS influences chromosome choreography and Ori positioning , we analyzed the number and localization of markers as a function of cell size in WT and maoP deletion cells . maoP inactivation reduced the number of cells containing 3–4 foci and increased the number cells with two foci for markers in the Ori region ( Fig 4C and S4 Fig ) . Newborn WT cells contain an Ori-3 marker that is rapidly duplicated and segregated to ¼ and ¾ positions of the cell . In contrast , in maoP cells , duplication of foci occurred later in larger cells and the average distance separating the two segregated foci was reduced in medium sized cells with positioning at the home position being less precise ( Fig 4C ) . In contrast , inactivation of maoP had no effect on the number and position of markers in Right and Ter MD ( S5 Fig ) . These results indicate that MaoP/maoS system affects the timing , positioning and separation of markers in the Ori MD . Two 150-kb regions in the Ori MD contain late splitting snap loci and the late separation of newly replicated DNA in these regions resulted from an interaction with SeqA [28 , 29] . Inactivation of MaoP had the opposite effect as it resulted in a slight delay in the separation of loci suggesting that Ori constraining by MaoP is independent of the late seqA-dependent segregation of snap regions ( S4 Fig ) . The molecular bases responsible for segregation of Ori to ¼–¾ positions following replication in E . coli are uncharacterized . The excision of DNA rings from Ori region revealed that two hours after excision , either no foci were evident in ~40% of cells or the focus was localized at one pole in ~60% of cells ( S1 Text and S6 Fig ) . These results suggest that upon excision , DNA rings accumulated at the pole and in the absence of replication were inherited linearly in the cell population . Events leading to the formation of cells with polar foci and cells without foci were visualized using time-lapse experiments ( S1 Text and S6 Fig ) . To test whether a large segment of Ori region was required to target the Ori region at the ¼- ¾ positions , we excised large segments from the Ori region , all lacking oriC . In each case , a parS tag targeted by ParB-GFP was inserted in the excisable segment carrying oriC and nucleoid DNA was visualized by DAPI staining . DNA rings composed of 500–600 kb segment excised from the Ori region behave similarly to DNA rings of 150 kb in that they were found located at one cell pole and did not co-localize with nucleoid ( Fig 5 ) . The same outcome was obtained with segments originating from either side of oriC . Combined , these results suggest that no simple determinant was identified in the ori region to specify its positioning at the ¼-¾ of the cell . To further explore how Ori regions segregate at ¼–¾ positions of the cell , we excised three additional segments of the Ori region carrying oriC . Two extended clockwise from the region close to oriC into the Ori MD whereas the third one extended counterclockwise from oriC towards the NSLeft region ( Fig 6A ) . Because the distance separating the two recombining sites is large ( > 500 kb ) , the rate of excision did not exceed 50% as estimated by the number of viable colonies . Excision gave rise to two circles , one 4 Mb remnant chromosome that is not replicated and one 500–600 kb “mini-chromosome” that is replicated via oriC . Excised rings and remnant chromosomes were visualized by parS/ParB-GFP , whilst nucleoid DNA was visualized by either DAPI staining or mCherry-HupA that uniformly labels cellular DNA . Two hours after the excision , two types of cells were detected . Firstly , a number of cells displayed a pattern of parS tags and nucleoids similar to cells in which excision was not induced ( Fig 6B , panel “no excision” ) . The number varied from experiment to experiment and presumably corresponded to cells that did not sustain an excision event . Secondly , cells contained two well separated masses of DNA , with multiple parS foci specifically localized with one DNA mass while the other mass was devoid of any parS focus ( Fig 6B , panel “excision” ) . The presence of multiple parS tags indicates that 600-kb rings carrying oriC were able to replicate . These results suggest that , at this time of observation , multiple replicated rings remained as a single nucleoid that was separated from the 4 Mb remnant chromosome found in the other half of the cell or in the adjacent new born cell . The use of strains carrying a parS tag in the mini-chromosome and either another FROS tag on the remnant chromosome ( either in Ori region or in Ter region ) or carrying a MatP-mCherry fusion protein confirmed the separation of mini-chromosomes carrying oriC from the remnant chromosome ( Fig 6C ) . The position of Ori markers suggested that rings carrying oriC separated at the edge of the nucleoid . The separation of circles carrying oriC from the 4-Mb non-replicating DNA was visualized by excising DNA segments in the presence of cephalexin to block cell division . Excised rings carrying Ori markers ( indicated by arrowheads ) were clearly separated from the 4-Mb non-replicating rings ( Fig 6D ) . Movies recapitulating the steps leading to the above situation were recorded using HU-labeled nucleoid and markers located in the excised rings carrying oriC ( Fig 6E ) . Images were taken every 15 minutes during a 4h period after recombinase induction . Seventy five minutes after excision , parS foci were separated from nucleoid DNA that was devoid of Ori markers ( indicated by red arrowheads ) ( Fig 6E ) . At subsequent times ( 105 , 120 and 150 min ) , the number of fluorescent foci as well as the amount of DNA localized with Ori markers increased . At times 120 min and 150 min , small nucleoid became apparent around Ori markers ( black arrowheads ) . Combined , these results suggest that the replication process and the expansion of nucleoid played a major role in chromosome positioning and segregation in the cell .
This current study presents the first description of a single cis acting site , maoS , and a trans acting protein , MaoP , involved in the organization of the Ori region in E . coli . Interestingly , MaoP belongs to a group of proteins involved in DNA metabolism and chromosome organization that coevolved with Dam methylase in Enterobacteria including MatP , MukBEF , SeqA , MetJ and other proteins of unknown function [30] . MatP constrains DNA over the 800-kb long Ter MD by binding to 23 matS target sites scattered throughout the Ter MD [7] . How a single maoS site constraints DNA over a distance of several hundred of kilobases is unknown , but may involve among other hypotheses a mechanism of tracking or loop extrusion ( for review , see [31] ) initiated at maoS to impose constraints over the large Ori MD . MaoP lacked homology to any characterized proteins , so how it structures the Ori MD remains unknown; furthermore , we cannot exclude the involvement of other components in this process ( es ) . The Ori MD has been identified by different approaches ( FISH , long-distance genetic interactions , segregation pattern and mobility of markers ) and has been shown to be preferentially targeted by retrotransposition of Ll . LtrB group II intron in slow growth conditions [32] . Probing the retrotransposition pattern of this element in a maoP mutant would reveal whether this preferential targeting is dependent of Ori constraining by MaoP . The insertion of segments originating from the Right MD impedes the effects of MaoP/maoS on distal segments . This could result from determinant ( s ) in the Right MD that block the process or the absence of determinant ( s ) in the Right MD that would be required for the process to be propagated further . As the insertion of segments from NS regions acquire Ori properties when inserted in the Ori MD , it is likely that segments from the Right MD contain determinants that antagonize the propagation of Ori properties . Further structure-function analysis is required to further characterize the factors and mechanisms constraining the Ori region and to uncover how segments from the Right MD might impede this process . DNA Rings of 150–200 kb excised from various chromosomal regions accumulated at the cell pole and remained localized in this cellular space for several generations , suggesting that they could not travel across the cellular territory occupied by the nucleoid . These observations are supported by the radial confinement of the nucleoid within the cell [11] and also suggest that 100 kb-long DNA polymers that are devoid of partition systems and unable to replicate do not mix with nucleoid DNA . The position of the excised ring at the time of excision was dependent on the original location in the genetic map . Rings excised from the Ori region were found preferentially at the old poles whereas rings from the Ter MD were found at midcell/new pole . Two models describing chromosome conformation and segregation have been proposed ( for review [2 , 33] ) . These models rely either on the folding of the chromosome as a random coil-like polymer compacted by external crowding forces and entropy to act as the main driver for chromosome segregation [34] , or on the presence of a folded chromosome , self adhering object with intrinsic structuring , interacting with various systems that control the conformation of different regions [8 , 11 , 13 , 35] . The presence of multiple replicating mini-chromosomes confined only in a small fraction of the cell is in favor of the second model . In E . coli , the nucleoid DNA does not occupy the full interior cell space whilst the space at the end contains aggregated proteins [36] and is accessible to F plasmids devoid of partition system [37] or other smaller plasmids [15 , 38 , 39] . Here , large non-replicating DNA molecules accumulated in these nucleoid free territories suggesting that an unidentified system controls central positioning of the nucleoid at midcell . In many bacteria , chromosome segregation relies on the highly conserved parABS system . In many cases , the centromere-like parS sequences are found near oriC suggesting a coordination of segregation with replication . In B . subtilis , the large Ori region including oriC and parS sequences forms a macrodomain and its 3D folding pattern plays a role in the regulation of replication initiation , chromosome organization and DNA segregation [40] . Interestingly , Enterobacteria lack parABS and no analogous system has been identified . The widespread conservation of parS across diverse bacteria suggests that Par systems evolved early in the evolution of bacterial chromosomes and were subsequently lost from the Enterobacteria . This loss may have resulted either from the acquisition of alternative segregation systems or the acquisition of new properties in DNA metabolism and chromosome management that rendered parABS dispensable . Here , to identify cis-elements required for chromosome segregation we used a strategy that , if they exist , should favor their positioning at the ¼ and ¾ positions . However , only excised segments carrying oriC localized to the center of the cell; furthermore , the replication process and the expansion of nucleoid played a major role in chromosome segregation and positioning . Interestingly , multiple copies of mini-chromosomes did not immediately segregate to other half of the cell or from each other . Instead they remained together in a growing bulk suggesting that DNA confinement may be important for driving chromosome segregation and/or that an interplay with cell cycle events and/or cellular structures are required for efficient chromosome segregation .
The bacterial strains and plasmids used in this study are listed in Tables 2–4 and in S1 Table . E . coli strains were grown at 30°C in Lennox broth ( rich medium ) , or in minimal medium A supplemented with 0 . 12% casaminoacids and 0 . 2% glucose . Antibiotics were added when necessary . The different transpositions targeted in the chromosome were performed as in [27] . The deletions and the insertions of specific segments were done by the one-step technique in strain DY330 and transduced into working strains as previously described [41 , 42] . Constructed strains were verified by PCR . Flow cytometry analyses were performed as described before [5] . Inversions tests were performed as described before , for 20 min at 36°C and for 10 min at 37°C [4] . To reduce variability of the assay in the maoP deletion background , the integrase-excisionase module was integrated in the chromosome using phage HK022-based integrative vectors [43] . The strains used for excision carried two attL and attR sites derived from the λ site-specific integration module . The two att sites were inserted in the same orientation . The induction of the collision was performed at 39°C for 20 minutes and the cells were subsequently incubated at 30°C for two hours . An aliquot of the cells was plated at the same time as the non-induced strain in order to estimate lethality . When the excised segments were about 200 kb long , the level of lethality , reflecting the amount of excision , was greater than 90% . When large segments of 500 kb—600 kb were involved , the level of excision was about 50–60% . To perform time lapses analyses , cells were spread immediately after induction on slides and analyzed under the microscope . For the cephalexin experiments , cephalexin ( 20 μg/ml ) was added 10 minutes before the induction of the excision . Cultures were grown in minimal A medium in the presence of glucose and casaminoacids without IPTG to maintain a minimal level expression of gfp-parB present on plasmid pALA2705 . For the observation of nucleoids , the cells carried a mCherry fusion upstream the hupA gene . Cells were plated on an agarose pad in the same medium and immediately observed under the microscope . For the short time lapse experiments , movies were recorded automatically on a Leica microscope . Autofocus was performed at every time point on the phase contrast image and GFP fluorescence was recorded on the plane with the best phase contrast focus . Image analysis was performed with ImageJ software using the manual tracking plugin ( http://rsb . info . nih . gov/ij/index . html ) . The XY co-ordinates of the two poles and of the foci were recorded manually and processed automatically with Excel ( Microsoft ) software . The travelled distance ( given in μm ) was estimated over a period of 5 minutes by adding up the absolute values of the distances for all 10 sec interval as described before [5] . The ( x , y ) coordinates of the foci at every time point were recorded and the distance travelled in the 10 s interval calculated . For the mobility measurements at the home position , 30 foci were analyzed ( 15 cells with two foci for markers ) . To measure the mobility of markers , the distance travelled by various foci was recorded when markers were at home position . In the growth conditions used , at home position , markers of Ori , Right MDs as well as markers from NS regions were segregated in the two halves of the cells .
|
The Ori region from bacterial chromosomes plays a pivotal role in chromosome organization and segregation as it is replicated and segregated early in cell division cycle and its positioning impacts the cellular organization of the chromosome in the cell . The E . coli chromosome is divided into four macrodomains ( MD ) defined as large regions in which DNA interactions occurred preferentially . Here we have identified a new system responsible for specifying properties to the Ori MD . This system is composed of two elements: a cis-acting target sequence called maoS and a gene of unknown function acting in trans called maoP . Remarkably , MaoP belongs to a group of proteins conserved only in Enterobacteria that coevolved with the Dam DNA methylase and that includes the MatP protein structuring the Ter macrodomain and the SeqA and MukBEF proteins involved in the control of chromosome conformation and segregation . These results reveal the presence of a dedicated set of factors required in chromosome management in enterobacteria that might compensate , at least partially , for the absence of the ParABS system involved in the condensation and/or segregation of the Ori region in most bacteria .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"bacteriology",
"medicine",
"and",
"health",
"sciences",
"chromosome",
"structure",
"and",
"function",
"microbiology",
"chromosome",
"mapping",
"dna",
"replication",
"bacterial",
"genetics",
"forms",
"of",
"dna",
"dna",
"molecular",
"biology",
"techniques",
"microbial",
"genetics",
"microbial",
"genomics",
"research",
"and",
"analysis",
"methods",
"bacterial",
"genomics",
"musculoskeletal",
"system",
"gene",
"mapping",
"chromosome",
"biology",
"chromosomal",
"dna",
"molecular",
"biology",
"biochemistry",
"cell",
"biology",
"nucleic",
"acids",
"anatomy",
"genetics",
"biology",
"and",
"life",
"sciences",
"genomics",
"chromosomes"
] |
2016
|
The MaoP/maoS Site-Specific System Organizes the Ori Region of the E. coli Chromosome into a Macrodomain
|
In northern Tunisia , the co-circulation of two related sand fly-borne phleboviruses , Toscana virus ( TOSV ) and Punique virus ( PUNV ) was previously demonstrated . In contrast to TOSV , a prominent human pathogen , there is no data supporting that PUNV is capable to infect and cause disease to humans . We studied the respective involvement of TOSV and PUNV in human infections in northern Tunisia through a seroprevalence study . The presence of TOSV and PUNV neutralising antibodies ( NT-Ab ) was tested in human sera collected from 5 districts of the governorate of Bizerte , and the titres of NT-Ab were estimated by microneutralisation ( MN ) assay . A total of 1 , 273 sera were processed . TOSV and PUNV NT-Ab were detected in 522 ( 41% ) and 111 sera ( 8 . 72% ) respectively . TOSV seroprevalence varied from 17 . 2% to 59 . 4% depending on the district . Analysis of TOSV geometric mean titre values demonstrated a constant increase according to the age . The vast majority of sera containing NT-Ab were found to be more reactive toward TOSV than PUNV . Indeed , past infections with PUNV and TOSV were undisputable for 5 and 414 sera , respectively . PUNV may be capable to infect humans but at a low rate . TOSV is responsible for the vast majority of human infections by sand fly-borne phleboviruses in northern Tunisia . TOSV must be considered by physician and tested in diagnostic laboratories for patients with meningitis and unexplained fever in northern Tunisia .
The risk of human infection with sand fly-transmitted viruses has been shown to cover extended geographic areas ( southern Europe , Africa , Middle-East , central and western Asia ) because of the presence of the sand fly vectors [1] . In countries bordering the Mediterranean basin , phlebotomine sand flies are involved in the transmission of several arthropod-borne viruses that belong to the genus Phlebovirus within the Bunyaviridae family . These sand fly-borne phleboviruses belong to three distinct serocomplexes : ( i ) the Sandfly fever Naples virus serocomplex including Toscana virus ( TOSV ) and related viruses ( Naples , Tehran , Massilia , Granada , Punique… ) , ( ii ) the Sandfly fever Sicilian virus serocomplex including Sicilian virus and related viruses ( Cyprus , Turkey… ) , and ( iii ) the Salehabad virus serocomplex including Salehabad virus and related viruses ( Arbia , Adria… ) [2] . Several of those viruses are recognised human pathogens ( TOSV , Naples virus , Sicilian virus , Cyprus virus and Adria virus ) [1] , [3] , [4] , [5] . Recent studies ( case reports , seroprevalence studies and virus isolation ) indicate that TOSV circulates actively in the Mediterranean area . TOSV is the only sand fly-borne phlebovirus which has been undoubtedly identified as an aetiological agent of neuroinvasive infections such as meningitis , meningo-encephalitis or peripheral neurological manifestations [6] , [7] , [8] . In Northern Mediterranean countries , infections due to TOSV represent an important public health problem as it is one of the major viral pathogens involved in aseptic meningitis during the warm season , i . e . between April and October [9] , [10] , [11] . Recent discoveries of new sand fly-borne phleboviruses from Mediterranean countries has indicated that the viral diversity in genus the Phlebovirus is higher than initially suspected [12] , [13] , [14] , [15] . In Tunisia , the recent isolation a new phlebovirus named Punique virus ( PUNV ) , from phlebotomine sand flies collected in the north of the country raised the question of its potential implication as a human pathogen [13] . Indeed , PUNV is antigenically and genetically closely related to but distinct from TOSV , and subsequently , it was included in the species Sandfly fever Naples virus . However , there is currently no additional evidence suggesting that PUNV is capable to infect and cause disease to humans . Recently , TOSV was suspected to circulate in Tunisia based on serological study [16] . These serological findings were corroborated by the recent isolation of TOSV from sand flies collected in Northern Tunisia [17] . The objective of the present work was to evaluate and to compare the respective involvement of TOSV and PUNV through a seroprevalence study in human , among a population at risk for sand fly-transmitted diseases originated from Northern Tunisia , by using the two viruses in a comparative manner through a microneutralisation ( MN ) assay .
Sera were collected from February to April , 2011 , from out care patients visiting local hospitals for medical reasons that were not available to us and requiring blood analysis . These patients are originated from 5 districts ( Mateur , Utique , Joumine , Sejenane and Ras Jabel ) of the governorate of Bizerte , Northern Tunisia located in the vicinity of the site where TOSV and PUNV were isolated repeatedly from sand flies in 2008 , 2009 , 2010 ( Figure 1 ) [13] , [17] . This study was performed with leftovers of these samples . The tubes were anonymized and only the sex , age and district address were recorded . This study was approved by the ethical committees of the Pasteur Institute of Tunis under the agreement number IPT/UESV/19/2010 , and of the Marseille Federation of Research No 48 under the number 13-008 . The virus microneutralisation ( MN ) assay , previously described for phleboviruses [18] , was adapted with minor modifications . Briefly , MN assay was performed in 96-well microtitre plates using Vero cells ( ATCC CCL81 ) . Two-fold serial dilutions from 1∶10 to 1∶80 were prepared for each serum and a volume of 50 µL was pipeted into 96-well plates , using an epMotion 5075 working station ( Eppendorf ) . The two virus strains were ( i ) Toscana virus strain MRS2010-4319501 ( GenBank accession nos KC776214–KC776216 ) isolated from a human case of meningitis in Southeastern France in 2010 [19] , and ( ii ) Punique virus T101 isolated from Phlebotomus sp . in Tunisia in 2009 ( Strain Tunisie2009T101 ) . The two virus strains were titrated in Vero cells . A volume of 50 µL containing 100 TCID50 was added into each well except for the controls that consisted of PBS . The plate containing 100 TCID50 of virus and the four two-fold dilutions ( 1∶10 to 1∶80 ) of serum was incubated at 37°C for one hour . Then , a 50 µL suspension of Vero cells containing approximately 2 . 105 cells in 5% foetal bovine serum was added to each well , and incubated at 37°C in presence of 5% CO2 . After 5 days , the microplates were read under an inverted microscope , and the presence or absence of cytopathic effect was noted . The titre ( no neutralisation , neutralisation at 1∶10 , 1∶20 , 1∶40 and 1∶80 ) was recorded . The threshold for positivity was defined as 1∶20 . Differences in titres lower than four-fold dilutions were considered as not significantly different . Serum exhibiting paired results such as neg/≥1∶20 , 1∶10/1∶20 , 1∶10/≥1∶40 , and 1∶20/≥1∶80 were indicative of a single infection against the virus corresponding to the highest dilution . Serum exhibiting paired results such as neg/1∶10 , were considered as negative for both viruses . Serum exhibiting paired results such as 1∶20/1∶40 , and 1∶40/1∶80 were indicative of past infection with both viruses . The GMT observed in MN with TOSV and PUNV were calculated respectively . Sera exhibiting an absence of neutralisation were attributed a score of 5 . Sera exhibiting neutralising properties were attributed the reciprocal of the dilution ( 10 , 20 , 40 or 80 ) . Dilutions ≥1∶160 were not tested since long range analysis ( 1∶10 to 1∶2560 ) of 100 randomly sorted sera indicated that titres ≥160 were seldom observed .
A total of 1 , 273 sera ( corresponding to 345 men and 928 women , sex ratio 0 . 37 ) were collected . The median age was 53 years ( range: 2–97 ) . They consisted of 86 , 484 , 244 , 240 , and 219 sera collected from the districts of Jounine , Mateur , Ras Jabel , Sejenane , and Utique , respectively ( Figure 1 ) . Detailed characteristics of the tested sera are presented in Table 1 . Neutralising antibodies against TOSV ( TOSV NT-Ab ) were detected in a total of 522 sera ( 41% ) : 96 had titre 10 , 116 had titre 20 , 165 had titre 40 , and 145 had titre 80 ( Table 2 ) . Neutralising antibodies against PUNV ( PUNV NT-Ab ) were detected in a total of 111 sera ( 8 . 72% ) : 99 had titre 10 , 11 had titre 20 , 0 had titre 40 and 1 had titre 80 ( Table 2 ) . Results are presented in Table 2 and detailed analysis is given as Text S1 . According to a 1∶20 cut-off for positivity and four-fold dilutions of difference , only five sera ( bolded values in table 2 ) reflected indisputable infection by PUNV , and a total of 414 sera ( stared values in table 2 ) possessed TOSV NT-Ab demonstrating infection by TOSV . For 144 sera ( underlined values in table 2 ) , possible cross-neutralisation between PUNV NT-Ab and TOSV NT-Ab precluded definitive interpretation and conclusion . These results suggested that the presence of TOSV NT-Ab may be responsible for PUNV cross-neutralisation . As shown in Table 3 , PUNV MN titres are tightly correlated with previous immunisation against TOSV , due to cross-neutralisation . In contrast , TOSV MN titres are poorly impacted by PUNV MN GMT , suggesting that the presence of TOSV MN NT-Ab can be , in a large majority of cases , unequivocally attributed to TOSV infection ( Table 4 ) . Detailed results of TOSV are presented globally for the 1 , 273 sera and for each region individually in Table 5 , respectively . At titre 10 , 41% of sera contained antibodies capable to neutralise TOSV . Among all districts of the Governorate of Bizerte , seroprevalence rates varied from 17 . 2% to 59 . 4% . The lowest seroprevalence rates were observed in the Ras Jabel district; the two districts of Sejenane and Joumine exhibited intermediate rates ( 30% and 40 . 7% ) ; the highest rates were observed in the districts of Mateur and Utique ( 50 . 2% and 59 . 4% ) . The proportions of sera capable to neutralise TOSV at titre 10 were maintained when analysis was performed at titres 20 , 40 or 80 . GMT analysis by age group is presented in Figure 2 . In Utique and Joumine districts , the number of individuals in group 0–20 was too small to be compared with other age groups . So those two points were not drawn in Figure 2 . However , those individuals were taken into account for analysis of the global population . Analysis of GMT values on the global population ( irrespective to the district ) demonstrated a constant increase according to the age . Similar trends were observed in the Mateur , Sejenane and Ras Jabel regions , independantly . In Utique and Joumine regions , the GMT values were stable and decreased in the oldest group . These differences with global population and other districts are not due to the size of this age group; it might be due to a bias in the tested population which is not representative of the global population .
In the Mediterranean area , several phleboviruses are circulating as demonstrated by virus isolation and/or molecular detection in sand flies , and some of them ( e . g . TOSV , Naples virus and Sicilian virus ) are recognised human pathogens [20] . TOSV is the leading cause of CNS infection in Southern European countries [9] , [11] . Interestingly , the US military medical literature reported the occurrence of sandfly fever in Northern Tunisia , namely in the regions of Tunis , Ferryville , Mateur and Bizerte during WWII in the US forces stationed in North Africa during the summer of 1943 [21] . The recent discovery of novel sand fly-borne phleboviruses ( Massilia virus , Granada virus , Punique virus ) that are antigenically and genetically closely related but clearly distinct from TOSV demonstrated that at least two of these viruses can co-circulate in a same geographic area [14] , [17] . These findings call for further investigation to elucidate the potential effect of these newly discovered phleboviruses on human health in these areas . Although it is known that TOSV can infect humans and cause a variety of clinical syndromes including neuro-invasive diseases , there is no or very limited data about the capacity of these newly discovered viruses to infect humans and to cause diseases . In Tunisia , PUNV strains have been isolated from Phlebotomus pernicisosus and Phlebotomus longicuspis collected from the district of Utique where TOSV strains have been also isolated [13] , [17] . Therefore , the demonstration of co-circulation questioned their respective role ( if any ) in human infections due to sand fly-borne phleboviruses in Northern Tunisia . Both viruses belong to the same virus species , Sandfly fever Naples , and consequently it is difficult to distinguish between them by using broadly reactive serological tests , such as inhibition hemagglutination assay , complement fixation assay , enzyme-linked immunosorbent assay ( ELISA ) or indirect immunofluorescence assay [6] , [13] , [18] , [22] , [23] , [24] . Indeed , serological cross-reactivity in a function of viral antigenic closeness: the more similar the viruses , the more cross-reactive the antibodies . The recent report of the presence of IgM and IgG reactive against TOSV using ELISA test indicates that either TOSV or an antigenic relative ( such as PUNV ) is involved in human infection in Northern Tunisia [16] , [25] . However , the lack of discrimination of ELISA cannot solve the problem of cross-reactivity and thus cannot indisputably involve TOSV as the etiologic agent of the CNS infections . The growing evidence that distinct but antigenically related sand fly-borne phleboviruses circulate in certain countries such as Spain ( TOSV and Granada virus ) , France ( TOSV and Massilia virus ) , and Tunisia ( TOSV and PUNV ) [9] , [10] , [13] , [14] , [15] , [17] pointed out to the cross-reactivity by using ELISA , IFAT and subsequently lead to conducting studies using neutralisation test which are the only assay with suitable discriminative capacity [6] , [18] , [26] . To attempt the determination of the respective role of TOSV and PUNV in human infection , a sero-epidemiological study concerning a population living in endemic areas for visceral leishmaniasis originated from Northern Tunisia was performed . A total of 1 , 273 sera were tested using MN assay , with the two viruses independently . In agreement with other studies [27] , [28] , we determined an “a priori” cut-off value at titre 20 , and analysed our results according to the observed MN titre . The vast majority of sera containing NT-Ab were found to be more reactive toward TOSV than to PUNV . Previous infection by PUNV or a closely related antigenic variant was undisputable for 5 sera . By contrast , previous infection by TOSV was undisputable for 414 sera . This demonstrates that although the two viruses are present in sand fly populations , TOSV is involved at a much higher frequency in human infection than PUNV . Interestingly , virus studies conducted on sand flies trapped in the same regions suggested that TOSV circulates at lower level than PUNV , since the latter was detected and isolated 6 times versus 2 times for TOSV of a total of 8 , 206 sand flies trapped during 3 successive seasons from 2008 to 2010 [13] , [17] . Our results indicate that PUNV ( or closely related antigenic variants ) can infect humans , but it occurs seldom in a region where the virus circulates at high level in sand fly populations . It should be underlined that this does mean that PUNV is not capable to cause human disease , but only that it is involved at a much lower rate in human infections than TOSV in this region of Tunisia . The clinical presentation associated with PUNV ( mild , similar or drastically different from TOSV infection ) in humans is currently unknown . Thus , its possible medical interest deserves further investigations ( e . g . , by investigating summertime undetermined febrile illness in the regions where the virus circulates ) . Seroprevalence rates of TOSV NT-Ab observed in this study ( global rate 41% , 17 . 2% to 59 . 4% depending on the district ) are much higher than those ( 2–25% ) reported in countries of southern Europe such as Portugal , Spain , France , Italy , Greece and Turkey [27] , [29] , [30] , [31] , [32] , [33] . Only few studies reported a seroprevalence higher than the one observed in Tunisia: in the Tuscany region , a seroprevalence of TOSV of 77 . 2% was reported among a population at high-risk ( forestry workers ) [32]; in Greece seroprevalence rates ranging from 39% to 51 . 7% were reported in several islands [34] . Although this study was not performed with a panel of sera representative of the population , it suggests that TOSV circulates at much higher frequency than in southern Europe . Similar studies should be performed in other regions of Tunisia , but also in other North African countries to better characterize this trend . Analysis of GMT values on the global population demonstrated a constant increase according to the age . Similar result were reported in other studies concerning various populations from endemic countries , where both anti-TOSV seroreactivity and TOSV-specific antibody prevalence increased significantly with age [27] , [34] , [35] . This trend indicates that TOSV infection can occur at any age of the life , and that repeated infections could play a role in sustained and increasing immunity as reflected by neutralising antibodies . These results deserve further confirmation by studies addressing much larger populations covering wider geographic areas . In conclusion , this study conducted in Northern Tunisia showed: ( i ) TOSV is responsible for the vast majority of human infections by sand fly-borne phleboviruses , ( ii ) PUNV , a recently discovered sand fly-transmitted phlebovirus that co-circulates with TOSV , is capable of infecting humans but at a low rate , ( iii ) important variations among seroprevalences are observed depending on the geographic area , and thus on environmental factors , and ( iv ) TOSV should be considered as an important pathogen and that needs to be included in all virological diagnostic concerning patients with meningitis and unexplained febrile illness originated from Northern Tunisia .
|
In northern Tunisia , two different pheboviruses are known to circulate in sand fly population , Toscana virus ( TOSV ) and Punique virus ( PUNV ) . In contrast to TOSV , a prominent human pathogen , there is no data supporting that PUNV is capable to infect humans and to cause a disease . We studied the respective involvement of TOSV and PUNV in human infections in northern Tunisia through a seroprevalence study . Because TOSV and PUNV are antigenically and genetically closely related , it is difficult to distinguish between them by using broadly reactive serological tests , such as enzyme-linked immunosorbent assay ( ELISA ) . Thus , we developed a method of microneutralisation assay using the two viruses in a comparative manner . A total of 1 , 273 sera were processed . We provide first evidence to support ( i ) that Punique virus may be capable to infect humans but at a low rate , ( ii ) that TOSV , the most prevalent arbovirus in Southern Europe , is responsible for the vast majority of human infections by sand fly-borne phleboviruses in northern Tunisia . Therefore , it is important to consider TOSV as an important pathogen that needs to be included in all virological diagnostic concerning patients with meningitis and unexplained febrile illness originated from Northern Tunisia .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[] |
2013
|
Co-Circulation of Toscana Virus and Punique Virus in Northern Tunisia: A Microneutralisation-Based Seroprevalence Study
|
No ideal vaccine exists to control plague , a deadly dangerous disease caused by Yersinia pestis . In this context , we cloned , expressed and purified recombinant F1 , LcrV antigens of Y . pestis and heat shock protein70 ( HSP70 ) domain II of M . tuberculosis in E . coli . To evaluate the protective potential of each purified protein alone or in combination , Balb/C mice were immunized . Humoral and cell mediated immune responses were evaluated . Immunized animals were challenged with 100 LD50 of Y . pestis via intra-peritoneal route . Vaccine candidates i . e . , F1 and LcrV generated highly significant titres of anti-F1 and anti-LcrV IgG antibodies . A significant difference was noticed in the expression level of IL-2 , IFN-γ and TNF-α in splenocytes of immunized animals . Significantly increased percentages of CD4+ and CD8+ T cells producing IFN-γ in spleen of vaccinated animals were observed in comparison to control group by flow cytometric analysis . We investigated whether the F1 , LcrV and HSP70 ( II ) antigens alone or in combination can effectively protect immunized animals from any histopathological changes . Signs of histopathological lesions noticed in lung , liver , kidney and spleen of immunized animals on 3rd day post challenge whereas no lesions in animals that survived to day 20 post-infection were observed . Immunohistochemistry showed bacteria in lung , liver , spleen and kidney on 3rd day post-infection whereas no bacteria was observed on day 20 post-infection in surviving animals in LcrV , LcrV+HSP70 ( II ) , F1+LcrV , and F1+LcrV+HSP70 ( II ) vaccinated groups . A significant difference was observed in the expression of IL-2 , IFN-γ , TNF-α , and CD4+/CD8+ T cells secreting IFN-γ in the F1+LcrV+HSP70 ( II ) vaccinated group in comparison to the F1+LcrV vaccinated group . Three combinations that included LcrV+HSP70 ( II ) , F1+LcrV or F1+LcrV+HSP70 ( II ) provided 100% protection , whereas LcrV alone provided only 75% protection . These findings suggest that HSP70 ( II ) of M . tuberculosis can be a potent immunomodulator for F1 and LcrV containing vaccine candidates against plague .
Plague caused by Y . pestis ( a Gram negative bacterium ) is a zoonotic infectious disease that has profoundly affected the course of history [1] , [2] and troubles human populations , leading to millions of deaths . According to the World Health Organization ( WHO ) , plague has been classified as a re-emerging infectious disease [3] . Rodents are the reservoirs for Y . pestis and the fleas transmit the bacteria from rodent to rodent . Infected fleas also transmit bubonic plague , the most common form of the disease from rodents to humans [4]–[6] . Humans are infected accidently after bites from fleas having Y . pestis , by direct contact with blood and tissues of infected animals , or by direct aerosol transmission . The aerosol transmission develops lethal pneumonic plague . The intentional aerosolization of Y . pestis in human population is the main concern of bioterrorism [7] . Plague can be treated with antibiotics at early stage . It has been reported that antibiotic-resistant strains of Y . pestis bacilli have been isolated in Madagascar and Mongolia [8] , [9] and showed naturally acquired multi-drug-resistant variants of Y . pestis [10] . These studies suggest that there is an urgent need to develop an effective vaccine that can provide long term protection and to counter the drug resistant variants of Y . pestis . Administration of live attenuated Y . pestis vaccine provides protection against plague in animal models [11] , [12] . These live attenuated plague vaccines are accessible in some countries , like Russia [13]; however , in the United States and Europe , these vaccines have never been licensed most probably due to several risk factors associated with the use of live-attenuated or whole cell killed vaccine in terms of side effects and administration of numerous antigens from live/killed vaccines [13]–[16] . Hence it is very much essential to develop new generation vaccines . Earlier studies using F1/LcrV-based vaccines that protect mouse models and cynomolgus macaques against aerosolized Y . pestis but they confer poor and inconsistent protection in African Green monkey models [17] , [18] . Further in order to improve the efficacy of F1/LcrV-based vaccines , several approaches are in progress . Amongst these , genetically modified antigens [19] , use of alternate adjuvants [20] , [21] and delivery systems [22] , [23] are very important as these approaches are certainly promising . It is noteworthy to mention that F1-negative Y . pestis strains persists [24] , and LcrV variants of Y . pestis may pose serious challenge for any vaccine with respect to cross-protection [25] , [26] . With this background , one possible strategic approach could be the inclusion of additional antigen/s that might play the role of an immunomodulator/s or and an immunoregulator/s to augment the immune response in the subunit vaccine preparation to encounter the possible disease threat . It has been established in the recent studies that subunit vaccines protect mouse models by inducing F1/LcrV-specific humoral immune response; however , to achieve complete protection cell mediated immune response mainly relies on the type-1 cytokines i . e . , IFN-γ and TNF-α [27]–[29] . These findings suggest that the efficacy of subunit vaccines might be improved if they induce Y . pestis-specific IFN-γ and TNF-α secreting memory T cells in addition to F1/LcrV-specific humoral immunity . In this scenario , it would be highly valuable to modulate the immune response of F1/LcrV antigens to create an effective plague vaccine . In context to this , the heat shock proteins70 are well documented to augment the immune response for the development of vaccine initiatives [30]–[35] . It has been proven that the role of HSP70 ( II ) in stimulating effective T-cell responses [36] to pathogen-specific antigens . As reported earlier , HSP70 ( II ) of M . tuberculosis is known to play crucial role in antigen processing and presentation by MHCs [37] . Huang et al . [36] demonstrated the role of fusion construct using ovalbumin-HSP70 , domain II [38] , amino acid ( 161–370 ) of HSP70 from M . tuberculosis , is sufficient to elicit ovalbumin specific CD8+ cytotoxic T lymphocytes ( CTLs ) . Here , in order to evaluate the HSP70 ( II ) as an immunomodulator , we have cloned caf1 and lcrV genes of Y . pestis and hsp70 ( II ) gene of M . tuberculosis . The encoding proteins were expressed in E . coli and purified upto homogeneity . In order to evaluate the protective efficacy , Balb/C mice were immunized with purified proteins F1 , LcrV , and HSP70 ( II ) alone or in combinations . Humoral and cell mediated immune responses were also evaluated . Immunized animals were challenged with 100 LD50 of Y . pestis via intra-peritoneal route . Significantly high IgG response was observed in the sera of immunized mice with F1 and LcrV alone or in combinations . Three combinations i . e . , LcrV+HSP70 ( II ) , F1+LcrV and F1+LcrV+HSP70 ( II ) provided 100% protection . HSP70 ( II ) modulated cellular immune response as the significantly elevated levels of IL-2 , IFN-γ , TNF-α and IFN-γ secreting CD4+/CD8+ T cells were noticed in spleen of F1+LcrV+HSP70 ( II ) group in comparison to the F1+LcrV group . HSP70 ( II ) also increased protective efficacy of LcrV from 75 to 100% . We also performed the histopathological studies to examine the liver , spleen , lung and kidney tissues from immunized animal groups that were intraperitoneally infected with virulent Y . pestis at 3rd and 20th day post infection . Y . pestis localization in tissues was also examined by immunohistochemistry using fluorescent microscopy .
Institutional Animal Ethics Committee ( IAEC ) of Defence Research and Development Establishment “approved” all the protocols for experiments conducted using mice wide registration number 37/Go/C/1999/CPCSEA and Institutional Biosafety committee ( IBSC ) wide protocol no: IBSC/21/MB/UT/12 as per the institutional norms . The principles of good laboratory animal care were followed all through the experimental process . The mice were maintained in accordance with recommendations of committee for the purpose of control and supervision of experiments on animals , Govt . of India . A virulent strain of Y . pestis ( clinical isolate , designated as S1 ) recovered from a patient during a sporadic outbreak of primary pneumonic plague occurred in Northern India in 2002 [39] , [40] was used for challenging experiments . Frozen stock of Y . pestis was streaked on Brain Heart Infusion ( BHI ) agar plate and incubated at 28°C for 48 h . A single colony from BHI agar plate was further inoculated in 5 ml of BHI broth and grown at 28°C for 48 h and the colonies ( CFU/ml ) were counted . All live Y . pestis cultures and animal experiments were performed in BSL-3 facility , DRDE , Gwalior . E . coli host strain BL21 ( DE3 ) and DH5α were purchased from Invitrogen , USA . The expression vector pET 28a+ was from Novagen , USA . Y . pestis , S1 strain was grown in BHI broth at 28°C and the genomic DNA was isolated by DNeasy Blood and Tissue kit ( Qiagen , USA ) . The genomic DNA of M . tuberculosis was a generous gift from DFRL , Mysore , India . The genes caf1 and lcrV of Y . pestis and hsp70 ( II ) of M . tuberculosis were amplified by polymerase chain reaction ( PCR ) . The details of used oligos in this study are given in Table 1 . The individual amplicon was ligated into pET28a vector using compatible restriction sites . The individual ligated product was transformed into chemically competent cells of E . coli host strain DH5α and the positive clones were selected on Luria Bertani ( LB ) agar plates supplemented with kanamycin ( 50 µg/ml ) . The plasmid DNA was isolated by using QIAprep Spin Miniprep Kit ( Qiagen , USA ) from overnight grown culture corresponding to individual clone . In order to express the recombinant antigens , E . coli host strain BL21 ( DE3 ) cells were transformed with individual recombinant construct corresponding to caf1 , lcrV and hsp70 ( II ) . The positive transformants were selected on LB agar plates containing kanamycin ( 50 µg/ml ) and were inoculated into 5 ml of LB medium with kanamycin and grown at 37°C . Cultures at logarithmic phase ( OD600 ∼0 . 75 ) were induced with 1 mM isopropylthiogalactoside ( IPTG ) and grown for 3 h . The cultures were pelleted and the cells were lysed in sample buffer and analyzed by SDS–PAGE . The recombinant F1 was purified using Ni-NTA column ( Qiagen , USA ) under denaturing conditions using 8 M urea following our earlier standardized protocol [41] . Recombinant LcrV and HSP70 ( II ) were purified in native conditions using Ni-NTA column according to the manufacturer's instruction . The purity of the recombinant proteins was analysed by SDS-PAGE and confirmed through western blot using monoclonal antibodies specific for 6X-his tag ( Qiagen , USA ) . The purified proteins F1 , LcrV and HSP70 ( II ) were separated by SDS-PAGE and analysed by Western blot using hyper immune sera at 1∶1000 dilution . The purified proteins were dialyzed and concentrated by using Amicon ultra centrifugal filter devices ( Millipore ) and the concentrations were estimated by Bradford method [42] . The endotoxin levels were measured by Limulus Amoebocyte Lysates ( LAL ) QCL-1000 kit ( Cambrex Biosciences , USA ) as per the manufacturer's protocol . Immunogenicity of recombinant proteins alone or in combination and protection of immunized mice against virulent Y . pestis ( S1 strain ) was evaluated in 6–8 week old female Balb/C mice . The animals were taken in three batches and divided into 8 groups/batch ( 8 mice/group ) i . e . , Control group; HSP70 ( II ) group; F1 group; LcrV group; F1+HSP70 ( II ) group; LcrV+HSP70 ( II ) group; F1+LcrV group and F1+LcrV+HSP70 ( II ) group ( Figure 1d [A] ) . The animals of batch-I were used for evaluation of IgG antibody response and protection studies against Y . pestis challenge; batch-II for evaluation of cell mediated immune response ( cytokine profiling and the estimation of CD4+ and CD8+ T cells ) and batch-III for histopathological/immunohistochemical studies . All the animal groups were immunized subcutaneously with 10 µg/mouse of each purified corresponding antigen/s as designated by their group name in formulation with aluminium hydroxide gel ( 0 . 35% in sterile phosphate buffer saline , PBS ) . The animals of control group were injected with PBS only . The prime dose was given on day 0 followed by two boosters on day 14 and 21 . Blood was collected after first and second booster from each group on day 0 , 21 and 28 , sera were separated for IgG antibody response ( Figure 1d [B] ) . In order to determine the protective efficacy , all the immunized animals of batch-I were challenged with virulent Y . pestis ( S1 strain ) with 100 LD50 ( 1 LD50 = 103 CFU/mouse ) by intraperitoneal route on day 60 after the prime vaccination . The virulence and the LD50 of Y . pestis ( S1 strain ) have been characterized earlier by our group [40] . Survival of the animals was monitored for 30 days after challenge ( Figure 1d [B] ) . Infection was confirmed by isolation and growth of Y . pestis on blood agar plate from the different organs viz; lung , liver , spleen and kidney of dead animals . For histopathology , all the immunized animals of batch-III were challenged as described earlier for protection studies . At 3rd day post infection , three mice in each group were sacrificed and the organs viz: lung , liver , spleen and kidney were collected . The tissues were placed into 10% neutral buffered formalin , dehydrated in serial alcohol gradient ( 70 , 80 , 90 and 100% ) , cleared with xylene , infiltrated in wax ( Leica TP-1020 ) and embedded in paraffin [44] . Three animals from each survived group i . e . , LcrV; LcrV+HSP70 ( II ) ; F1+LcrV and F1+LcrV+HSP70 ( II ) on day 20 post infection and three naive control animals ( neither immunized nor challenged ) were sacrificed . As described above , their tissues were removed and fixed in 10% neutral buffered formalin for paraffin block preparation . Various sections of 4–5 µm thickness were prepared ( Microm HM-360 ) and stained with haematoxylin and eosin ( HE ) and analysed under light microscope ( Leica , DMLB ) . For the presence of Y . pestis , the tissues sections were also used for immunohistochemical studies [45] . Briefly , sections were deparaffinised , cleared with xylene and rehydrated . The tissues sections were washed with PBS and subjected to antigen-retrieval by boiling in 0 . 1 M citrate buffer [pH 6 . 0] for 10 min . The sections were then incubated with 3% H2O2 in methanol for 10 min to block the endogenous peroxidase activity and blocked with 5% skimmed milk in PBS for 2 h . The tissue sections were incubated with mouse anti-F1 antibody at 1: 1000 dilutions for overnight at 4°C . After three washings with PBS ( each for 5 min ) , sections were incubated with FITC-labelled rabbit anti mouse secondary antibody for 1 h at room temperature and again washed thrice with PBS . The tissue sections were cover slipped using PBS/glycerine ( 1∶3 ) and observed under fluorescence microscope ( Leica , Germany ) using Leica application suit software . Statistical comparisons for IgG titers , cytokine levels and IFN-γ secreting CD4+ and CD8+ T cells were performed . Analysis was done using SigmaStat 3 . 5 , by one way ANOVA , All Pairwise Multiple Comparison Procedure ( Fisher LSD Method ) . *P<0 . 05; ** P<0 . 01; *** P<0 . 001; # P<0 . 001 . Gehan-Breslow-Wilcoxon test was used to compare the protective potential against Y . pestis infection amongst different vaccinated . Survival curve analysis ( percentage survivals ) was done by Kaplan Meier's method ( ****P<0 . 0001 , ***P<0 . 001 ) . The genes caf1 , lcrV of Yersinia pestis and hsp70 ( II ) of M . tuberculosis were used in this study for primer designing under the NCBI accession AF074611 . 1 , NC003131 . 1 and CP002992 . 1 respectively . The gene sequences to lcrV and caf1 from Y . pestis ( S1 strain , an Indian clinical isolate ) were submitted to GenBank at NCBI under the Accession No . KF682423 and KF682424 respectively .
The genes caf1 ( 513 bp ) encoding F1 ( ∼17 kDa ) , lcrV ( 981 bp ) encoding LcrV ( ∼38 kDa ) of Y . pestis and hsp70 ( II ) ( 630 bp ) encoding a domain II of HSP70 ( ∼23 kDa ) of M . tuberculosis were cloned in the pET 28a vector . The in-frame and the orientation of the cloned genes were confirmed by nucleotide sequencing ( Chromous , Biotech , India ) . The schematic diagram ( Figure 1a ) of the three recombinant proteins represents the location of histidine tag and orientation of open reading frame . The nucleotide sequences to lcrV and caf1 genes from Y . pestis ( S1 strain , an Indian clinical isolate ) were submitted to GenBank at NCBI under the Accession No . KF682423 and KF682424 respectively . The recombinant constructs corresponding to F1 , LcrV and HSP70 ( II ) were transformed in BL-21 ( DE3 ) . Small-scale cultures of the positive clones were subjected to IPTG induction to identify clones capable of expressing the predicted size of recombinant proteins . The expression profile of recombinant proteins F1 , LcrV and HSP70 ( II ) were analysed by SDS-PAGE . A typical induction experiment comparing the polypeptide SDS-PAGE profiles of un-induced and IPTG-induced cultures for F1 , LcrV and HSP70 ( II ) are shown in Figure 1b [A] , [B] and [C] respectively . To facilitate the purification of the recombinant proteins , the constructs were designed to carry the 6X-His tag either at N-terminus or C-terminus . Lysis under native conditions revealed the association of recombinant F1 with the pellet fraction , demonstrating that the F1 protein was insoluble . However , LcrV and HSP70 ( II ) were associated with supernatant fractions , demonstrating that LcrV and HSP70 ( II ) were soluble . The purification of the LcrV and HSP70 ( II ) was carried out in native conditions , however , F1 carried out by solubilizing in 8 M urea and purified by Ni-NTA affinity chromatography . The purified recombinant proteins were analysed by SDS-PAGE as shown in Figure 1c . The proteins i . e . , F1 [A]; LcrV [B] and HSP70 ( II ) [C] observed to be almost pure . The concentrations of the purified proteins were estimated and the yield of F1 , LcrV and HSP70 ( II ) was 14 , 20 and 25 mg/L of shake flask cultures respectively . In a western blot experiment , anti-histidine antibody recognized these proteins corresponding to their molecular weights . Immunoblot with hyper immune sera against F1 , LcrV and HSP70 ( II ) recognized the corresponding proteins ( Figure S1 ) . The endotoxin content performed by LAL assay of purified protein was less than 5EU per 25 µg of each purified protein . To evaluate the IgG endpoint titers in all the vaccinated groups , total IgG were measured to F1 and LcrV in sera samples collected seven days after first and second boosters respectively . The cut-off value for the assays was calculated as the mean OD ( +2 SD ) from sera of control group assayed at 1∶100 dilution . The endpoint IgG titers were calculated as reciprocal of the highest serum dilution giving an OD more than the cut-off . In order to compare the protective efficacy , the immunized animals were challenged with 100 LD50 of virulent Y . pestis including control group . Survivals of the animals were monitored for 30 days post challenge ( Figure 6 ) . Three vaccine combinations [LcrV+HSP70 ( II ) , F1+HSP70 ( II ) , F1+LcrV+HSP70 ( II ) ] resulted in 100% protection from the Y . pestis challenged mice ( P<0 . 0001 ) , whereas the LcrV and F1+HSP70 ( II ) vaccinated mice were only 75% ( P<0 . 001 ) and 12 . 5% protected , respectively . There was no protection observed in control , HSP70 ( II ) and F1 groups . Y . pestis was recovered from the spleen , lung , liver and kidney of dead animals which succumbed to the challenge and identified by the growth on blood agar . Survived animals were sacrificed 30 days post-challenge , and autopsied for any bacterial presence in their organs like spleen , lung , liver and kidney . Vaccinated animals that survived the challenge appeared to clear Y . pestis from the mice since no growth was observed on blood agar plates from spleens , lungs , livers , and kidneys . On day 3 and 20 after challenge with virulent Y . pestis ( S1 strain ) , the lung , liver , kidney and spleen of the immunized groups including control group were isolated , fixed and prepared for HE staining . Normal mice that were neither immunized with plague vaccines or PBS nor infected with Y . pestis were used as naive controls . The animals sacrificed on day 3 after infection , histopathological lesions were observed in the lung tissues ( Figure 7a ) . Normal alveolar pattern with normal alveolar septa , air duct , alveoli and bronchioles with intact epithelium were observed from naive control group ( Figure 7a [A] ) whereas all the vaccinated including control group , lung parenchyma showed inflammation including neutrophil infiltration into the airways and alveoli as shown by arrow ( Figure 7a [B] ) . The significant lung lesions were congestion , hemorrhage , granulovacuolar degeneration of bronchiole associated lymphoid tissue , bronchial lumen occlusion and psuedomembrane formation ( Figure 7a [B-I] ) . Survived animals from LcrV; LcrV+HSP70 ( II ) ; F1+LcrV and F1+LcrV+HSP70 ( II ) vaccinated groups effectively recovered as no histopathological lesions were observed ( Figure 7a [J-M] ) . In spleen ( Figure 7b ) , normal architecture with white pulp consisting of lymphatic follicles and red pulp consisting of sinusoidal and other element of blood were observed from naive control mice ( Figure 7b [A] ) whereas all the vaccinated animals including control group showed reduced density of white pulp follicles and congestion in the red pulp , lymphoid follicle depletion ( arrow ) , lacking of lymphocytes , exhibiting higher number of myeloid and erythroid lineage cells and also presence of megakaryocytes as shown by bold arrow ( Figure 7b [B-I] ) . Survived animals from LcrV; LcrV+HSP70 ( II ) ; F1+LcrV and F1+LcrV+HSP70 ( II ) vaccinated groups showed regression of splenic lesions except LcrV group that offered less protection and few megakaryocytes were seen ( Figure 7b [J-M] ) . In kidney ( Figure 7c ) , normal glomerulus , Bowman's space and renal parenchyma were observed from naive control mice ( Figure 7c [A] ) whereas the vaccinated and control group showed parenchymal granular degeneration ( bold arrow ) , fragmentation of the chromatin material and renal tubule showing cloudy swelling with hydropic degeneration shown by arrow ( Figure 7c [B-I] ) . Survived animals from LcrV; LcrV+HSP70 ( II ) ; F1+LcrV and F1+LcrV+HSP70 ( II ) vaccinated groups restored the normal appearance of renal capsule , glomeruli and renal tubules ( Figure 7c [J-M] ) . In liver ( Figure 7d ) , normal hepatic cord arrangement , hepatic lobes and hepatocytes with normal hepatic parenchyma were observed in naive control mice ( Figure 7d [A] ) whereas vaccinated and control groups , liver histology exhibited granulovacuolar degeneration of hepatocytes ( arrow ) , perinuclear clumping of the cytoplasm and obliteration of the chromatin material , few periportal and intraparenchymal small aggregates of macrophages and neutophils were seen ( Figure 7d [B-I] ) . Survived animals from LcrV; LcrV+HSP70 ( II ) ; F1+LcrV and F1+LcrV+HSP70 ( II ) vaccinated groups recovered hepatic lesions , less infiltration of mononuclear cells and vacuolar degeneration ( Figure 7d [J-M] ) . To study the dissemination of Y . pestis from peritoneal cavity to various vital organs i . e . , lung , spleen , liver and kidney by immunohistochemistry were performed . The immunized animals including PBS control were sacrificed on day 3 after infection to localize Y . pestis by immunohistochemistry in lung , spleen , liver and kidney ( Figure 8 ) . No bacterium was observed in lung , liver , spleen and kidney isolated from the naive control group where as the clumping of Y . pestis was observed from all the vaccinated animals including control group by immunohistochemistry ( Figure 8 ) . On day 20 after infection , survived animals from LcrV; LcrV+HSP70 ( II ) ; F1+LcrV and F1+LcrV+HSP70 ( II ) groups were sacrificed for bacterial localization . No bacterial presence was observed in any of the survivors by immunohistochemistry ( Fig . 8 ) .
Y . pestis suppresses the host immune system in susceptible animal species , but the infection survived animals can effectively overcome the re-infection . This hints the possibility of developing effective vaccine that can boost the immune defense mechanisms against plague . Although intensive studies are in progress for several decades on plague [46] there is no safe and efficient vaccine till date . The F1/V based subunit vaccine candidate that evokes mainly humoral immune response , although has shown promising results in animal models , its efficacy in humans is not yet evaluated [47] . Further , the next-generation plague vaccines that are yet to be developed should also evoke cell-mediated immune response [28] . Humoral and cellular immunity potentially contribute to vaccine efficacy [48] . Humoral immunity relies upon production of antibodies by plasma B cells which effectively neutralizes extracellular pathogens while cellular immunity relies upon cytokine-producing capacities of T cells and is particularly effective in eradicating intracellular pathogens [49] . The protection evoked by cell-mediated immune response against intracellular pathogens mainly relies on Th1 type of immune response , considered by the development of pathogen-derived antigen specific IFN-γ and TNF-α secreting T cells [50] , [51] . The Y . pestis replicates in macrophages of host and has developed a competent mechanism for the depletion of the NK cells that finally decreasing IFN-γ expression . The IFN-γ suppression obliterates the inflammatory response that is responsible for development of adaptive immunity [52] . It has been proved that STAT4-deficient mice with low level of IFN-γ were showing inadequate protection against Y . pestis infection despite producing high IgG antibody titers [53] . These findings indicate that high IgG titers may not be sufficient for vaccine efficacy . In case of plague , to develop an effective vaccine should evoke both humoral as well as strong Th1 type of cellular immune responses . Th1 type of immunity can assist to evoke the humoral immune response and to produce the long term memory cells . In vivo experiments proved that the administration of IFN-γ and TNF-α provide protection to mice against virulent Y . pestis challenge [54] . These evidences suggest that cellular immunity priming Y . pestis antigen specific Th1 CD4+ T cell is important for protection against plague . It is quite evident from the earlier studies that heat shock proteins ( HSPs ) are known to elicit potent T-cell responses not only to model antigens [31] , [55] but also to the pathogen-derived antigens [35] , [56] . HSP70 ( II ) of M . tuberculosis is one of the examples to these various antigens , has been proven to evoke the T-cell response by several groups [31] , [35] , [55] . Ovalbumin-HSP70 ( II ) ( domain II ) fusion constructs elicit ovalbumin-specific CD8+ cytotoxic T lymphocytes [36] . It has been demonstrated by Suzue and Young in 1996 that HSP70 ( II ) of M . tuberculosis enhance the humoral and cellular immune response to the p24 protein of HIV1 [30] . In the present study , we evaluated three recombinant proteins F1 , LcrV from Y . pestis and HSP70 ( II ) ( domain II ) from M . tuberculosis . In order to augment the immune responses , HSP70 ( II ) was formulated with F1 and LcrV and the animals were immunized with different combinations of antigen/s in formulation with aluminium hydroxide gel , a human compatible adjuvant . Sera from mice immunized with LcrV; LcrV+HSP70 ( II ) ; F1+LcrV; F1+LcrV+HSP70 ( II ) group had higher LcrV-specific IgG titers in comparison to F1-specific IgG titers in F1; F1+HSP70 ( II ) ; F1+LcrV and F1+LcrV+HSP70 ( II ) groups . HSP70 ( II ) significantly induced high F1 and LcrV-specific serum IgG titers in F1+HSP70 ( II ) ; LcrV+HSP70 ( II ) and F1+LcrV+HSP70 ( II ) immunized groups in comparison to F1 , LcrV and F1+LcrV groups respectively . There are four IgG subclasses viz; IgG1 , IgG2a , IgG2b , and IgG3 to provide the immunity against most of the infectious agents . In cell-mediated immune response , there is a change in the predominant immunoglobulin class or classes of the specific antibody produced . T-cells and their cytokines are mainly responsible to control the switch of these isotypes . Th1 type of immune response signals via STAT4 to produce cytokines such as IFN-γ and IL-2 to favour a strong cellular immunity , whereas IL-4 signals via STAT-6 to favour a humoral immune response and thus biased towards Th2 type of immune response [53] . In this study , we observed significantly high level of Th1 type of cytokines i . e . , IL-2 , IFN-γ and TNF-α in the splenocytes from all the vaccinated groups upon in vitro stimulation with group specific antigen/s in comparison to control group . HSP70 ( II ) significantly modulated the expression level of IFN-γ in F1+HSP70 ( II ) ; LcrV+HSP70 ( II ) and F1+LcrV+HSP70 ( II ) immunized groups in comparison to F1 , LcrV and F1+LcrV groups respectively . In case of IL-2 , a significant difference was observed in LcrV+HSP70 ( II ) and F1+LcrV+HSP70 ( II ) in comparison to LcrV and F1+LcrV groups respectively whereas TNF-α was observed in F1+LcrV+HSP70 ( II ) group in comparison to F1+LcrV group . No significant difference in the expression level of Th2 type of cytokines ( IL-4 and IL-10 ) was noticed . CD4+ T cells play an important role in the development of cellular immune responses and maintenance of memory CD8+ T cell responses [57] . The roles for CD8+ T cells during Y . pestis infection is not yet clear , but Y . pestis maintains virulence in the host by suppressing the production of Th1 type of cytokines [58] . Here , IFN-γ secreting CD4+ and CD8+ T cells were enumerated by flow cytometric analysis . A significant difference was observed in IFN-γ secreting CD4+ and CD8+ T cells in all vaccinated groups in comparison to control group . HSP70 ( II ) significantly increased the IFN-γ secreting CD4+ and CD8+ T cells in F1+LcrV+HSP70 ( II ) immunized group in comparison to F1+LcrV group . Histopathological assessment is valuable for evaluating the efficacy of new plague vaccines and for better understanding of the pathogenesis of the disease progression . To investigate whether the F1 , LcrV and HSP70 ( II ) antigens alone or in combination can effectively protect immunized animals from any histopathological alterations . Signs of histopathological lesions were noticed in lung , liver , kidney and spleen of immunized animals on 3rd day post challenge . To examine the histopathological changes in survived animals of LcrV; LcrV+HSP70 ( II ) ; F1+LcrV and F1+LcrV+HSP70 ( II ) groups , three animals from each group were sacrificed on 20th day post infection . The survived animals did not display any histopathological lesions in all the examined tissues . Immunohistochemistry showed bacteria in lung , liver , spleen and kidney on 3rd day post infection whereas no bacterium was observed on 20th day post infection in survived animals of LcrV , LcrV+HSP70 ( II ) , F1+LcrV and F1+LcrV+HSP70 ( II ) vaccinated groups . Several lines of evidence suggest that the outer surface proteins F1 and LcrV of Y . pestis are considered as the leading vaccine candidates and have been formulated to develop a subunit plague vaccine in the recent past [59]–[61] , [48] . F1+LcrV combination can fully protect rodent models against lethal Y . pestis challenge [47] , [62] however these vaccines provide poor and inconsistent protection ( between 0 and 75% ) in African Green monkeys [16] . Although these antigens are poorly immunogenic however their immunogenicity could be enhanced in formulation with Alum adjuvant [58] or by making a fusion protein with a molecular adjuvant like flagellin [63] . In this study , F1 and LcrV antigens have been formulated with HSP70 ( II ) as an immunomodulator to augment the immune response of these two vaccine candidates . In mouse model , LcrV alone provided 75% protection whereas LcrV+HSP70 ( II ) formulation provided 100% protection . F1 alone completely failed to protect whereas F1+HSP70 ( II ) provided 12 . 5% protection . F1+LcrV and F1+LcrV+HSP70 ( II ) provided 100% protection . Our finding proved that HSP70 ( II ) enhanced the protective potential of F1 and LcrV vaccine candidates in mouse model however these formulations need to be tested in non human primates .
|
Efforts are in progress by various scientific groups towards the development of plague vaccines . However , lack of better understanding about the Y . pestis infection mechanisms and pathogenesis prevents the development of an effective vaccine . In our effort to develop a more efficacious plague vaccine , we evaluated the role of HSP70 ( domain II ) of M . tuberculosis in formulation with the F1 and LcrV subunits of Y . pestis vaccine candidates . It is well documented that the F1 and LcrV alone does not always provide complete protection whereas a mixture of the F1+LcrV provides 100% protection in mouse model but poorly protect African green monkey models . In this study , LcrV provided 100% protection in formulation with HSP70 ( II ) whereas LcrV alone could provide only 75% protection in Y . pestis challenged mice . Two another combinations i . e . , F1+LcrV and F1+LcrV+HSP70 ( II ) also provided 100% protection whereas HSP70 ( II ) or F1 alone failed to protect . HSP70 ( II ) also modulated cellular immune response as the significantly elevated levels of IL-2 , IFN-γ , TNF-α and IFN-γ secreting CD4+/CD8+ T cells were noticed in spleen of F1+LcrV+HSP70 ( II ) group in comparison to the F1+LcrV group . These findings describe the role of HSP70 ( II ) and propose future perspectives for development of new generation plague vaccine .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"bacteriology",
"immune",
"cells",
"immunology",
"microbiology",
"vaccines",
"infectious",
"disease",
"immunology",
"immunologic",
"adjuvants",
"vaccination",
"and",
"immunization",
"bacterial",
"genes",
"immunomodulation",
"animal",
"cells",
"molecular",
"biology",
"immune",
"system",
"immunopathology",
"prophylaxis",
"cell",
"biology",
"clinical",
"immunology",
"immunity",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"vaccine",
"development",
"lethality",
"(bacteriology)"
] |
2014
|
HSP70 Domain II of Mycobacterium tuberculosis Modulates Immune Response and Protective Potential of F1 and LcrV Antigens of Yersinia pestis in a Mouse Model
|
Tissue macrophages are derived exclusively from blood monocytes , which as monocyte-derived macrophages support HIV-1 replication . However , among human tissue macrophages only intestinal macrophages are non-permissive to HIV-1 , suggesting that the unique microenvironment in human intestinal mucosa renders lamina propria macrophages non-permissive to HIV-1 . We investigated this hypothesis using blood monocytes and intestinal extracellular matrix ( stroma ) -conditioned media ( S-CM ) to model the exposure of newly recruited monocytes and resident macrophages to lamina propria stroma , where the cells take up residence in the intestinal mucosa . Exposure of monocytes to S-CM blocked up-regulation of CD4 and CCR5 expression during monocyte differentiation into macrophages and inhibited productive HIV-1 infection in differentiated macrophages . Importantly , exposure of monocyte-derived macrophages simultaneously to S-CM and HIV-1 also inhibited viral replication , and sorted CD4+ intestinal macrophages , a proportion of which expressed CCR5+ , did not support HIV-1 replication , indicating that the non-permissiveness to HIV-1 was not due to reduced receptor expression alone . Consistent with this conclusion , S-CM also potently inhibited replication of HIV-1 pseudotyped with vesicular stomatitis virus glycoprotein , which provides CD4/CCR5-independent entry . Neutralization of TGF-β in S-CM and recombinant TGF-β studies showed that stromal TGF-β inhibited macrophage nuclear translocation of NF-κB and HIV-1 replication . Thus , the profound inability of intestinal macrophages to support productive HIV-1 infection is likely the consequence of microenvironmental down-regulation of macrophage HIV-1 receptor/coreceptor expression and NF-κB activation .
Macrophages play crucial roles in the establishment , pathogenesis and latency of human immunodeficiency virus-1 ( HIV-1 ) infection [1] , [2] , [3] through their ability to support viral replication [4] , [5] , transmit virus [6] and act as a viral reservoir [6] , [7] , [8] , [9] . In this connection , macrophages throughout the body , including lymphoid tissue macrophages [10] , [11] , brain microglia [12] and genital ( vaginal ) macrophages [13] , are permissive to HIV-1 . In sharp contrast , resident macrophages in the human small intestine are profoundly incapable of supporting productive HIV-1 infection [13] , [14] , [15] , although intestinal macrophages are derived exclusively from blood monocytes [16] , which when differentiated into monocyte-derived macrophages are HIV-1 permissive [4] , [5] , [17] , [18] . The unique non-permissiveness of intestinal macrophages to HIV-1 stands in marked contrast to the ability of intestinal CD4+ T cells to support productive viral infection and undergo early , rapid and profound depletion during primary HIV-1 and SIV infection [19] , [20] , [21] , [22] , [23] , [24] , [25] , [26] . After their recruitment into the lamina propria , pro-inflammatory blood monocytes differentiate into non-inflammatory intestinal macrophages through stromal transforming growth factor β ( TGF-β ) -mediated Smad-induced IκBα and nuclear factor kappa B ( NF-κB ) inactivation , as we recently reported [27] , [28] . In further contrast to blood monocytes , intestinal macrophages are markedly down-regulated for receptors that mediate inflammatory responses , including LPS , Fcγ and Fcα receptors [27] , [28] , [29] , triggering receptor expressed on myeloid cells-1 ( TREM-1 ) [30] , [31] , as well as CD4 , CCR5 and CXCR4 [13] , [14] , [15] . Since CCR5 expression correlates directly with the differentiation of monocytes into macrophages [32] , [33] , [34] , the reduced expression of CCR5 on intestinal macrophages raises the possibility that the non-permissiveness of intestinal macrophages to HIV-1 is related to reduced HIV-1 receptor/co-receptor expression . However , our detection of proviral DNA in isolated intestinal macrophages exposed to HIV-1 in vitro [14] suggests post-entry restriction also may be involved in the inability of intestinal macrophages to support HIV-1 replication . To elucidate the mechanism that renders intestinal macrophages non-permissive to HIV-1 , we exposed blood monocytes and monocyte-derived macrophages to conditioned media from cultured lamina propria stroma isolated from normal human jejunum to determine the effect of the lamina propria microenvironment on CD4/CCR5 expression and HIV-1 permissiveness . Our results indicate that the inability of primary human intestinal macrophages to support HIV-1 replication is likely due not only to the marked down-regulation of CD4 and CCR5 but also to the inability of intestinal macrophages to activate NF-κB , a critical requirement for HIV-1 transcription .
CCR5-tropic HIV-1 strains are predominant among the transmitted/founder viruses isolated from acutely infected persons [35] , [36] , [37] . Since the gastrointestinal mucosa is the largest reservoir of macrophages in the body [38] , and macrophages are an important HIV-1 target cell , we initiated studies to define the HIV-1 receptor phenotype and permissiveness of purified intestinal macrophages to macrophage-tropic HIV-1 . Intestinal macrophages and blood monocytes were isolated from the same donors , purified and analyzed for expression of the HIV-1 primary receptor CD4 and the coreceptors CCR5 and CXCR4 . As shown in Table 1 , very low proportions of intestinal macrophages expressed CD4 ( 1 . 0% ) , CCR5 ( 0 . 8% ) and CXCR4 ( 2 . 1% ) , and a barely detectable proportion ( 0 . 3% ) expressed both CD4 and CCR5 ( P = 0 . 0001 to P = 0 . 039 ) , consistent with our earlier finding of markedly diminished CD4 , CCR5 and CXCR4 expression on intestinal but not vaginal macrophages [13] . The low levels of CD4 and CCR5 expressed on intestinal macrophages corresponded to low levels of receptor/co-receptor-specific mRNA [13] . In contrast , modest proportions of blood monocytes expressed CD4 ( 11 . 6% ) , CCR5 ( 2 . 9% ) and CXCR4 ( 14 . 1% ) , and 2 . 2% of the monocytes were CD4+CCR5+ , indicating that 3- to 10-fold fewer intestinal macrophages expressed the receptors compared to autologous blood monocytes ( Table 1 ) . We previously showed that isolated intestinal macrophages do not support HIV-1 replication [13] , [14] , [15] . The low level of CD4 , as well as CCR5 , on intestinal macrophages ( Table 1 ) raised the possibility that a restriction in HIV-1 entry could contribute to the cells' non-permissiveness to HIV-1 . To address this possibility , we sorted autologous CD4+ intestinal macrophages and blood monocytes by magnetic activated cell sorting ( MACS ) , cultured the cells for 4 days ( >98% viable ) , inoculated each population with equivalent amounts of highly fusigenic and macrophage-tropic R5 viruses , including NA420 B33 , NA20 B59 or NA353 B27 , which infect cells with extremely low levels of CD4 and/or CCR5 expression [39] , and monitored viral replication by p24 release over 20 days . As shown in Figure 1 , 95% of both the intestinal macrophages and blood monocytes were HLA-DR+CD13+ . Among the sorted CD4+ intestinal macrophages , 34 . 1% expressed CCR5 and levels of p24 were barely detectable only on day 12 , whereas among the sorted CD4+ blood monocytes , 26% expressed CCR5 and large amounts of p24 were released by the monocyte-derived macrophages up to day 20 ( Figure 1 ) . Importantly , neither the exposure of intestinal macrophages to pro-inflammatory stimuli , including lipopolysacharride , interferon-γ or tumor necrosis factor-α , nor culture for up to 2 weeks prior to inoculation with virus , induced HIV-1 permissiveness in the macrophages ( data not shown ) . These findings indicate that even CD4+ intestinal macrophages that express CCR5 are refractory to HIV-1 , implicating a post-entry mechanism for down-regulated HIV-1 permissiveness . However , the profound low level of CD4 and CCR5 expression on the total intestinal macrophage population ( Table 1 ) raised the possibility that the mucosal microenvironment of the jejunum caused the down-regulation of CD4 and CCR5 , thereby also contributing to the reduced permissiveness of intestinal macrophages to CCR5-tropic HIV-1 . Intestinal macrophages are terminally differentiated and express very low levels of CD4 and CCR5 , but they are derived from blood monocytes [16] , which , during and after differentiation into adherent macrophages , express high levels of CD4 and CCR5 . Since factors released by the intestinal extracellular matrix ( stroma ) down-regulate an array of innate response receptors on blood monocytes [27] , we examined whether stromal factors present in conditioned media derived from normal intestinal stroma ( S-CM ) [27] , [28] also down-regulate CD4 and CCR5 expression on blood monocytes during and after their differentiation into macrophages . Compared to monocytes differentiated into adherent macrophages during 2 days culture in media alone , monocytes differentiated into macrophages in the presence of S-CM ( 10–500 µg protein/mL ) displayed a marked dose-dependent decrease in surface CD4 and CCR5 ( Figure 2A ) . In contrast , when monocytes were first differentiated for 4 days into adherent macrophages and then exposed for 2 days to S-CM , CD4 and CCR5 expression was not down-regulated ( Figure 2B ) . Thus , intestinal stromal products prevent differentiation-induced upregulation of CD4 and CCR5 expression on monocyte-derived macrophages but do not down-regulate receptor/co-receptor expression after the cells have differentiated into macrophages . These findings offer an explanation for the near absence of CD4 and CCR5 on terminally differentiated intestinal macrophages , which are derived exclusively from circulating monocytes that have recruited into the lamina propria . Since undifferentiated monocytes do not support productive HIV-1 infection , we next determined whether monocyte-derived macrophages exposed to lamina propria stromal products supported HIV-1 replication . Monocyte-derived macrophages were cultured for 2 days in the presence of varying concentrations of S-CM , after which the cultures were inoculated with R5 virus ( NA353 B27 ) . As shown in Figure 3A , the pre-incubation of monocyte-derived macrophages with S-CM prior to the inoculation of HIV-1 caused a dose-dependent decrease in p24 production during a 20-day culture period . However , when monocyte-derived macrophages were pre-incubated with conditioned media from purified cultures of intestinal epithelial cells ( EC-CM ) [40] or intestinal mononuclear cells ( MNL-CM ) [27] derived from the same donor tissue as the S-CM , HIV-1 replication was not inhibited ( Figure 3B ) . Furthermore , S-CM also caused a dose-dependent decrease in viral replication when S-CM and virus were added simultaneously to the monocyte-derived macrophage cultures ( Figure 3C ) . These findings suggest that extracellular matrix products , rather than intestinal epithelial cell or lamina propria mononuclear cell products , inhibit productive HIV-1 infection in intestinal macrophages and that the down-regulation in viral replication is not the exclusive consequence of the low level of CD4 and CCR5 expression on the macrophages . To further distinguish between reduced HIV-1 entry and down-regulated viral replication , we pseudotyped HIV-1 with VSV-G envelope to bypass HIV-1 receptor/co-receptor-dependent entry . As predicted , treatment of monocyte-derived macrophages with S-CM for up to 24 hours did not impair the entry of VSV-G pseudotyped virus into the cells ( data not shown ) but caused a dose-dependent reduction in single-round replication of VSV-G pseudovirons , as shown by immunofluorescence and flow cytometry in Figure 4 , upper panels . The same pre-treatment of monocyte-derived macrophages with S-CM also inhibited infection of YU2 pseudovirons in a dose-dependent manner ( Figure 4 , lower panels ) . These results further indicate that S-CM inhibition of R5 replication was not due only to down-regulated CD4 and CCR5 expression but also involved post-entry restriction in viral replication . We have shown that stromal TGF-β inactivates NF-κB in monocyte-derived macrophages by dysregulating NF-κB signal proteins and inducing IκBα , the cytoplasmic negative regulator of NF-κB [28] . Because NF-κB is required for HIV-1 transcription [41] , we investigated whether stromal TGF-β-mediated down-regulation of NF-κB also inhibits the ability of monocyte-derived macrophages to support HIV-1 replication . Monocyte-derived macrophages were cultured in triplicate with increasing concentrations of S-CM and inoculated with R5 HIV-1 ( NA353 B27 ) at a multiplicity of infection ( MOI ) of 1 . After 2 hours , cells were visualized by confocal microscopy for the translocation of phosphorylated NF-κB p65 ( pNF-κB p65 ) into the nucleus and the cytoplasmic and nuclear intensity of NF-κB . On day 12 , the supernatants in parallel cultures were analyzed for the level of p24 . As shown in Figure 5A , exposure of monocyte-derived macrophages to increasing concentrations of S-CM caused a dose-dependent decrease in NF-κB p65 translocation into the nucleus and a dose-dependent decrease in p24 production . However , when S-CM at an inhibitory concentration of 250 µg protein/mL was pre-incubated for 1 hour with anti-TGF-β antibodies at a concentration of 100 µg/mL , S-CM inhibition of both the nuclear translocation of NF-κB p65 and HIV-1 p24 production was reversed , whereas pre-incubation with irrelevant IgG ( 100 µg/mL ) antibody had no effect on S-CM inhibitory activities ( Figure 5B ) . Furthermore , incubation of the cells with activated , recombinant human TGF-β ( rhTGF-β at a concentration of 10 pg/mL had little or minimal effect on NF-κB translocation or p24 production ( Figure 5C ) . However , rhTGF-β 50 pg/mL , which approximates the concentration of TGF-β in S-CM 250 µg/mL , inhibited NF-κB translocation and activity , as well as p24 production , similar to that of S-CM 250 µg/mL ( Figure 5C ) . Moreover , we previously showed ( flow cytometry , ELISA , immunocytochemistry and Western blot ) that LPS-exposed intestinal macrophages and S-CM-treated blood monocytes did not phosphorylate p65 , had very low levels of p50 , did not translocate p50 or p65 into the nucleus and expressed markedly reduced levels of NF-κB signal proteins ( 28 ) . Expression of p50 and p65 genes also were markedly reduced in intestinal macrophages compared to autologous blood monocytes ( 28 ) . These findings are consistent with minimal , if any , transcriptionally active p50/p65 heterodimer and together implicate stromal TGF-β-mediated down-regulation of NF-κB activation in the inhibition of HIV-1 replication by stromal factor-differentiated macrophages in vitro and intestinal macrophages in vivo .
We have shown that macrophages isolated from normal human small intestine are highly refractory to productive HIV-1 infection [13] , [14] , [15] , supporting observations that memory CD4+ T cells rather than macrophages are the predominant mononuclear target cell in the intestinal mucosa during primary HIV-1 infection [19] , [20] , [21] , [22] , [23] , [24] , [25] , [26] . We also have shown that in contrast to intestinal macrophages , vaginal macrophages are permissive to macrophage-tropic HIV-1 [13] . Since tissue macrophages throughout the body are derived from blood monocytes , our findings suggest that the lamina propria of the intestinal mucosa is a unique microenvironment capable of influencing HIV-1 permissiveness in blood monocytes recruited to the intestinal mucosa . Consistent with this concept , we present new evidence that products released by the intestinal extracellular matrix inhibit up-regulation of CD4 and CCR5 during the differentiation of blood monocytes into macrophages . However , the low level of CD4 and CCR5 expression on intestinal macrophages is not the exclusive cause of the cells' non-permissiveness to HIV-1 , since ( 1 ) the very small subset ( 1% ) of intestinal macrophages that express CD4 , a proportion of which also express CCR5 , did not support HIV-1 replication; ( 2 ) intestinal stromal products also decreased HIV-1 replication when stromal products were added simultaneously to cultures of monocyte-derived macrophages , i . e . , before the induction of CD4 and CCR5 down-regulation; and ( 3 ) stromal products inhibited single-round gene expression of VSV-G pseudotyped virus , which enters cells independent of CD4 and CCR5 . In this connection , we previously showed that unsorted intestinal macrophages with undetectable CD4 also do not of support HIV-1 replication ( 13 , 14 ) . Having previously shown that stromal TGF-β differentiates pro-inflammatory blood monocytes into non-inflammatory cells with the phenotype and function of intestinal macrophages [27] through Smad-induced IκBα expression and NF-κB signal dysregulation [28] , we show here that a critical consequence of stromal TGF-β-induced NF-κB inactivation is the profound inability of monocyte-derived macrophages to support HIV-1 replication . TGF-β is reported to both inhibit and stimulate HIV-1 replication , depending on the cell type , level of cell differentiation , virus strain , timing of treatment and presence of other cytokines [42] , [43] , [44] . In intestinal mucosa , latent TGF-β is produced by many different types of cells , including epithelial cells , mast cells , T regulatory cells , T cells undergoing apoptosis , and stromal cells . TGF-β constitutively released by these cells binds to the lamina propria extracellular matrix binding domains and upon activation and release regulates multiple macrophage defense and immune functions , consistent with an elaborate and finely tuned system of cross-talk that we have described previously [16] . Here we show that among these functions is the down-regulation of NF-κB activity and thus HIV-1 replication in monocyte-derived macrophages . These data suggest that TGF-β , at least in part , mediates the profound non-permissiveness of intestinal macrophages to HIV-1 . NF-κB plays a critical role in HIV-1 replication in T cells [41] and cells of the monocyte lineage [45] . In addition to stimulating the initiation of HIV-1 transcription [46] , [47] , [48] , NF-κB also has been implicated in promoting HIV-1 transcriptional elongation [49] , [50] . Importantly , NF-κB is constitutively activated in HIV-1-infected monocytes [51] , possibly through upstream activation of the IKK complex by HIV-1 regulatory/accessory proteins [52] , [53] or HIV-1-induced ( via NF-κB activation ) cytokines [54] . The activation of IKK leads to the phosphorylation and proteosomal degradation of IκBα and IκBβ , thereby releasing NF-κB for translocation into the nucleus to bind NF-κB-binding sites in the enhancer region of the HIV-1 long terminal repeat and host gene promotor sites . Thus , we conclude that stromal TGF-β inactivates NF-κB in monocyte-derived macrophages and that this inactivation likely contributes to the profound blockade in HIV-1 expression in intestinal macrophages , a highly unique population of mononuclear phagocytes [55] , [56] . The HIV-1 non-permissiveness of intestinal macrophages due to NF-κB inactivation is consistent with our recent finding that stromal TGF-β dysregulation of NF-κB signaling causes inflammation anergy in intestinal macrophages [28] . Importantly , long-term culture of intestinal macrophages in the absence of stromal factors does not restore inflammatory capability [27] , [28] and , as reported here , did not promote the emergence of HIV-1 permissiveness , indicating prolonged , if not permanent , down-regulation of these functions in intestinal macrophages . Also , exposure of intestinal macrophages to pro-inflammatory stimuli , including lipopolysaccharide ( LPS ) , interferon-γ ( IFN-γ ) and tumor necrosis factor-α ( TNF-α ) , does not induce inflammatory function [27] , [28] and did not restore replication competence . These findings suggest that in primary HIV-1 infection , resident macrophages in healthy intestinal mucosa are incapable of de novo HIV-1 replication . In contrast to primary HIV-1 infection , in late stage disease HIV-1-infected blood monocytes may recruit to intestinal mucosa that is either inflamed or infected with opportunistic pathogens . In such a microenvironment , dysregulated homeostasis permits viral replication to continue after the monocytes take up residence in the lamina propria , as we have reported for esophageal macrophages in patients with AIDS and opportunistic mucosal infections [57] . We also have reported that cytomegalovirus blocks stromal inhibition of HIV-1 infection of macrophages and that this inhibition is mediated , at least in part , by cytomegalovirus-induced monocyte production of TNF-α , which acts in trans to enhance HIV-1 replication [58] . However , the very low levels of TNF-α ( <2 . 9 pg/mL ) in S-CM generated from normal mucosa and inflamed Crohn's mucosa [59] suggest that TNF-α is not involved in stromal down-regulation of intestinal macrophage permissiveness to HIV-1 . In the present study , we investigated HIV-1 permissiveness in intestinal macrophages using highly macrophage-tropic R5 viruses , including NA420 B33 , NA20 B59 and NA353 B27 [39] , [59] , in order to maximize the possibility of infecting intestinal macrophages . Interestingly , infectious molecular clones of transmitted founder viruses derived from acutely infected persons are R5-tropic but fail to replicate efficiently in monocyte-derived macrophages [60] , [61] . Although we have not yet examined the ability of these molecular clones to infect intestinal macrophages , such infection seems unlikely , since intestinal macrophages do not activate NF-κB , a requirement for HIV-1 gene transcription during macrophage differentiation [45] . The findings presented here do not exclude the possibility that HIV-1 restriction factors other than TGF-β are present in the stroma and thus S-CM . S-CM was used in a range of 10-1000 µg protein/mL , corresponding to TGF-β in the range of 1–150 pg/mL . Although rhTGF-β at a concentration of 10 pg/L had little or minimal effect on NF-κB translocation or p24 production ( Figure 5C ) , rhTGF-β 50 µg/mL , which approximates the concentration of TGF-β in S-CM 250 µg/mL , inhibited NF-κB translocation and viral replication ( p24 production ) , similar to that of S-CM 250 µg/mL . Apolipoprotein B mRNA-editing enzyme catalytic polypeptide-like 3G ( APOBEC3G ) , which causes dC-to-dU mutations in viral DNA , is reported to be induced by LPS in dendritic cells and by IFN-α in monocyte-derived macrophages [62] , [63]; however , we have been unable to detect APOBEC3G in resting or IFN-α-treated intestinal macrophages . Also , higher levels of anti-HIV-1 miRNAs have been reported to inhibit HIV-1 in monocytes [64] , [65] , but the role of miRNA as a restriction factor in monocytes is controversial [66] , [67] . A cellular restriction factor that is neutralized by primate lentiviral Vpx protein was recently detected in quiescent monocytes , but its reduction as the cells differentiate into macrophages makes it an unlikely restriction factor in terminally differentiated intestinal macrophages [68] . Other potential restriction factors , including p21 [69] , [70] and interferon-induced C/EBPβ [71] , [72] , have been proposed but have not yet been investigated in mucosal macrophages . A confounding issue regarding post-entry restrictions in intestinal macrophages is that such restrictions would be unique to macrophages residing in the intestinal mucosa , since macrophages in the vaginal mucosa are highly replication competent [13] . Although the extracellular matrix could release products that induce yet-to-be-identified anti-viral restrictions , the findings presented here implicate stromal TGF-β-induced NF-κB inactivation as contributing to the non-permissiveness of macrophages in the human small intestine . These findings help explain the overwhelming absence of productive infection in intestinal macrophages , in sharp contrast to the highly productive infection in intestinal T cells , in primary HIV-1 infection . The ability of intestinal CD4+ T cells to support robust HIV-1 replication is well established in our in vitro [13] and in vivo studies [9] , [10] , [11] , [12] , [13] , [14] , [15] , [16] , [17] , [18] , [19] , [20] , [21] , [22] , [23] , [24] , [25] , [26] . Furthermore , TGF-β does not inhibit HIV-1 expression in a chronically infected T cell line or in primary T cell blasts infected in vitro with HIV-1 [42] . The discordance between intestinal T cell and macrophage support for HIV-1 replication in the presence of down-regulatory stromal TGF-β is currently under investigation in our laboratory . Thus , the unique dysregulation in NF-κB signaling induced in monocytes by extracellular matrix products , especially TGF-β , when the cells take up residence in the intestinal mucosa , offers a mechanism by which the host down-regulates mucosal macrophages for harmful pro-inflammatory responses and permissiveness to viruses in which transcription is NF-κB-dependent . Harnessing this natural anti-viral defense mechanism may provide a novel strategy to exploit for the prevention of infection in HIV-1 permissive cells .
All tissue and cell protocols were approved by the Institutional Review Board of the University of Alabama at Birmingham . Written informed consent was provided by study participants . Macrophages were isolated from segments of intestinal mucosa of otherwise healthy subjects undergoing elective gastric bypass by enzyme digestion and purified by counterflow centrifugal elutriation , as previously described [73] , [74] , [75] . Circulating blood monocytes from the same donors were purified by gradient sedimentation followed by magnetic anti-PE bead isolation of anti-CD14-PE-treated cells per the manufacture's manual ( Miltenyi Biotec ) . All studies were performed using fresh cells . Macrophages and monocytes were routinely >98% pure and 98% viable by propidium iodide staining . CD4+ monocytes and intestinal macrophages were isolated by magnetic CD4+ microbead separation . Macrophage-tropic viruses were prepared as previously described [13] , [76] , [77] . Briefly , replication competent clones of highly macrophage-tropic R5 viruses , including NA420 B33 , NA20 B59 and NA353 B27 [39] , were transfected into 293T cells by Fugene 6 ( Roche ) , according to the manufacture's protocol . After 60 hours , the supernatants were harvested , clarified by low speed centrifugation ( 1 , 000 g , 10 minutes ) , filtered ( 0 . 45 µm filter ) , titrated using JC53BL cells [78] , aliquoted and stored at −80°C . YU2 envelope ( Env ) or vesicular stomatitis virus glycoprotein ( VSV-G ) HIV-1 pseudovirions that express GFP upon infection were kindly provided by D . Levy , NYU and constructed as follows . Briefly , the env gene was deleted and the gfp gene was inserted between the env and nef genes of the pNL4-3 clone . An internal ribosome entry site ( IRES ) element was inserted between the gfp and nef genes to rescue nef gene expression [79] . To generate the YU2 Env or VSV-G GFP reporter pseudovirions , the clone was co-transfected with the YU2 Env or VSV-G expression plasmid into 293T cells and harvested , as described above . Using our previously described protocols [40] , [73] , [74] , epithelium and lamina propria mononuclear cells ( MNLs ) were removed by enzyme digestion from segments of normal human jejunum from otherwise healthy subjects undergoing elective gastric bypass , and purified by elutriation . The epithelial cells ( EC ) ( 10×106/mL ) , lamina propria MNLs ( 10×106/mL ) , and cell-depleted lamina propria stroma ( 1 g wet wt stromal tissue/mL ) , respectively , were cultured in RPMI for 24 hours without serum , and the EC-conditioned media ( EC-CM ) , MNL-CM and stroma-CM ( S-CM ) were harvested , sterile-filtered ( 0 . 2 mm Syringe Filter; Corning Inc . ) and frozen at −70°C , as previously described [27] , [28] . Cell depletion from lamina propria stroma was confirmed by immunohistochemistry [73]; intestinal macrophages expressed barely detectable CD14 [13] . Conditioned media did not alter monocyte-derived macrophage viability during incubation for as long as 4 days as assessed by flow cytometric analysis of propidium iodide uptake . S-CMs were normalized to 500 µg/mL RPMI . Endotoxin and protein content were determined by ELISA ( endotoxin ELISA: Cambrex Bio Science; protein ELISA: Pierce Protein Research Products/Thermo Scientific ) . Only endotoxin-free EC-CM , MNL-CM and S-CM were used in the experiments . Intestinal macrophages and monocytes were incubated with optimal concentrations of PE- , APC- , or FITC-conjugated antibodies to HLA-DR , CD13 , CD4 , CCR5 ( BD Pharmingen ) , or control mAbs of the same isotype at 4°C for 20 minutes , washed with PBS , fixed with 1% paraformaldehyde and analyzed by flow cytometry . Data were analyzed with FlowJo software ( Tree Star , Inc . ) . To examine the effect of S-CM on CD4 and CCR5 expression in monocyte-derived macrophages , blood monocytes were cultured in 48-well plates at 5×105 cells/well in RPMI plus macrophage colony-stimulating factor ( M-CSF ) serum and S-CM at final concentrations of 0 , 10 , 100 and 500 µg/mL for up to 3 days and analyzed for CD4 and CCR5 . Student's t-test was used to determine the statistical significance of the difference of expression levels of these receptors between intestinal macrophages and autologous blood . Sorted intestinal macrophages and monocytes from 2 donors were cultured in triplicate in 96-well plates at 2×105 cells/well in RPMI plus M-CSF and serum for 4 days . Cultures then were inoculated with NA20 B59 , NA353 B27 or NA420 B33 at an MOI = 1 , cultured for the indicated duration with 100 µL of supernatant , harvested every 4 days and stored at −70°C until assayed for p24 by ELISA ( PerkinElmer ) . To examine the effect of S-CM on macrophage permissiveness to HIV-1 , MACS-sorted monocytes were cultured for 4 days in RPMI plus M-CSF to generate monocyte-derived macrophages , after which S-CM was added at final concentrations of 10 , 100 and 500 µg protein/mL . Control cultures of monocyte-derived macrophages were incubated in media alone . Two days later , culture supernatants were removed , and triplicate cultures were inoculated with NA353 B27 ( MOI = 1 ) for 2 hours , cultured for 20 days , and the kinetics of p24 production was determined as above . Parallel triplicate cultures of monocyte-derived macrophages were inoculated simultaneously with NA353 B27 ( MOI = 1 ) plus S-CM ( final concentrations of 0 , 10 , 100 and 500 µg protein/mL ) for 2 hours , and viral replication was monitored as above . Cultures of monocyte-derived macrophages prepared as above were inoculated with NA353 B27 ( MOI = 1 ) plus S-CM or with S-CM only . Cells treated with S-CM only were harvested after 2 hours , cytospun onto glass slides and stained for NF-κB p65 . Cells infected with virus were cultured , and supernatants were harvested on day 12 and assayed for p24 by ELISA . Parallel monocyte-derived macrophages were inoculated for 2 hours in triplicate with NA353 B27 ( MOI = 1 ) plus S-CM 250 µg protein/mL pre-treated with 0 , 25 or 100 µg/mL of anti- TGF-β for 1 hour at 37°C . Analysis of viral replication and NF-κB p65 staining were performed as above . A final aliquot of monocyte-derived macrophages prepared as above was cultured for 6 days , inoculated in triplicate with NA353 B27 ( MOI = 1 ) plus rhTGF-β ( R&D Systems ) or rhTGF-β only at final concentrations of 0 , 10 , or 50 pg/mL for 2 hours . Evaluation of NF-κB p65 intensity and viral replication were performed as above . Cells cytospun onto glass slides were fixed and permeabilized with Cytofix/Cytoperm ( BD Biosciences ) for 20 minutes . After washing with PBS , cells were blocked with casein protein ( DAKO ) for 1 hour and incubated with rabbit anti-NF-κB p65 or isotype control antibodies ( Santa Cruz Biotechnology ) for 90 minutes , washed with PBS , incubated with donkey anti-rabbit IgG-FITC ( Jackson ImmunoResearch Laboratories ) for 30 minutes , washed with PBS and counterstained with DAPI nuclear stain . Cells were visualized by confocal microscopy , and the cytoplasmic and nuclear fluorescence intensity of NF-κB was converted to histograms using IPLab image analysis software version 3 . 6 ( BD Biosciences Bioimaging ) . For comparison of the effects of treatment on NK-κB activity , NF-κB intensity was normalized to the blue signal in the nucleus . Five images were analyzed per sample and mean intensities were generated . For comparison of the effects of treatment on HIV-1 replication , p24 value of each treatment was normalized to the media control group with the replication level of the media control group defined as 100% . Statistical significance was determined by Student's t-test . Data is expressed as mean ± SD or ± SEM , and statistical significance between groups was determined using Student's t-test . P values ≤0 . 05 were considered significant .
|
Human intestinal macrophages , unlike lymphoid tissue macrophages , brain microglia and genital ( vaginal ) macrophages , are profoundly incapable of supporting productive HIV-1 infection . Intriguingly , all macrophages are derived exclusively from blood monocytes , which are HIV-1 permissive after differentiation into monocyte-derived macrophages . Therefore , the unique non-permissiveness of intestinal macrophages to HIV-1 must be conferred by the intestinal mucosal microenvironment . Here we report that intestinal stroma potently blocked up-regulation of HIV-1 receptor/coreceptor CD4 and CCR5 expression during monocyte differentiation into macrophages and macrophage nuclear translocation of NF-κB , which is a critical requirement for HIV-1 transcription . These two mechanisms work collaboratively to render intestinal macrophages non-permissive to HIV-1 . Harnessing this natural antiviral defense may provide a novel strategy to exploit for the prevention of infection in HIV-1 permissive cells .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"infectious",
"diseases",
"virology",
"hiv",
"biology",
"microbiology",
"viral",
"diseases"
] |
2011
|
Stromal Down-Regulation of Macrophage CD4/CCR5 Expression and NF-κB Activation Mediates HIV-1 Non-Permissiveness in Intestinal Macrophages
|
Vascular endothelial growth factor ( VEGF ) is a potent cytokine that binds to specific receptors on the endothelial cells lining blood vessels . The signaling cascade triggered eventually leads to the formation of new capillaries , a process called angiogenesis . Distributions of VEGF receptors and VEGF ligands are therefore crucial determinants of angiogenic events and , to our knowledge , no quantification of abluminal vs . luminal receptors has been performed . We formulate a molecular-based compartment model to investigate the VEGF distribution in blood and tissue in humans and show that such quantification would lead to new insights on angiogenesis and VEGF-dependent diseases . Our multiscale model includes two major isoforms of VEGF ( VEGF121 and VEGF165 ) , as well as their receptors ( VEGFR1 and VEGFR2 ) and the non-signaling co-receptor neuropilin-1 ( NRP1 ) . VEGF can be transported between tissue and blood via transendothelial permeability and the lymphatics . VEGF receptors are located on both the luminal and abluminal sides of the endothelial cells . In this study , we analyze the effects of the VEGF receptor localization on the endothelial cells as well as of the lymphatic transport . We show that the VEGF distribution is affected by the luminal receptor density . We predict that the receptor signaling occurs mostly on the abluminal endothelial surface , assuming that VEGF is secreted by parenchymal cells . However , for a low abluminal but high luminal receptor density , VEGF binds predominantly to VEGFR1 on the abluminal surface and VEGFR2 on the luminal surface . Such findings would be pertinent to pathological conditions and therapies related to VEGF receptor imbalance and overexpression on the endothelial cells and will hopefully encourage experimental receptor quantification for both luminal and abluminal surfaces on endothelial cells .
Physiologic angiogenesis , the growth of new capillaries from pre-existing blood vessels , occurs in wound healing , pregnancy , exercise , and embryonic development . Diseases such as cancer and age-related macular degeneration are angiogenesis-dependent [1] . The growth of new capillaries from pre-existing blood vessels is mediated by several growth factors , one of which is a potent family of cytokines called vascular endothelial growth factor ( VEGF ) . The VEGF family is composed of five members: VEGF-A ( often referred to as VEGF ) , VEGF-B , VEGF-C , VEGF-D and placental growth factor ( PlGF ) . Alternative splicing of VEGF-A provides about 13 different VEGF isoforms [2] , [3] . Human VEGF consists of at least seven isoforms: VEGF121 , VEGF145 , VEGF148 , VEGF165 , VEGF183 , VEGF189 , and VEGF206 [4] , [5] . Although VEGF121 , VEGF165 , VEGF183 are diffusible , VEGF189 and VEGF206 are mainly sequestered in the extracellular matrix [4] . Amongst the major isoforms ( with length 121 , 165 , 189 and 206 amino acids ) , VEGF121 and VEGF165 are more highly expressed than VEGF189 and VEGF206 . Furthermore , the roles of VEGF189 and VEGF206 in vivo remain to be clearly identified [3] . For these reasons , we only consider VEGF121 and VEGF165 isoforms in the present model . These two isoforms bind VEGF receptors , VEGFR1 ( fms-related tyrosine kinase 1 or Flt-1 in humans ) and VEGFR2 ( kinase insert domain receptor also designated as Flk-1 , or KDR in humans ) . VEGF165 binds to the non-signaling co-receptor neuropilin-1 ( NRP1 ) as well and serves as a bridge for the VEGFR2-NRP1 complex . It has been shown recently that VEGF121 may also bind to NRP1; however , this binding is not sufficient to bridge the VEGFR2-NRP1 complex [6] . Preliminary sensitivity analyses from our group suggest that incorporation of the binding between VEGF121 and NRP1 does not drastically change the predictions regarding the VEGF distribution [7] . Therefore , this binding is not included in the model at the moment; this can be modified when more information becomes available . Finally , VEGF165 contains a heparin binding domain , which allows it to bind to the heparan sulfate glycosaminoglycan ( GAG ) chains of the extracellular matrix and the cellular basement membranes [8] . We have introduced a compartment model of VEGF distribution in the human body [9] . In the “healthy” set-up , the system was composed of two main compartments: the blood ( vascular system ) and the rest of the body . A third compartment was added for pathological cases to distinguish the diseased from the healthy tissue . VEGF121 , VEGF165 , and their respective interactions with VEGFR1 , VEGFR2 and NRP1 were considered . VEGF was secreted by the parenchymal cells ( in the healthy tissue ) and the tumor cells ( when the diseased tissue was assumed to be a breast cancer tumor ) . Other elements in the blood , such as platelets and granulocytes , sequester large amounts of VEGF and could potentially release significant amounts of VEGF as well [10]; the role of these processes in VEGF balance in the body is not known . However , since the rates of VEGF release from these blood elements have not been quantified , we have decided , as a first approximation , to neglect explicit representation of these sources; a distinct mathematical term can be added to the equations to model VEGF release from these elements in the future . We assume that the compartments are well-mixed and that freely diffusible ( unbound ) VEGF is transported by vascular permeability between the tissues and the blood . The model presented here is an extension of our previously published model [9] , as was a recent study that analyzed the effects of soluble VEGFR1 [7] , [11] . Two major additions were made . First , lymphatic drainage of VEGF was added , serving as a second route for VEGF to be transported from the tissue to the blood compartment . Secondly , our previous model considered VEGF receptors to be solely expressed on the abluminal endothelial surface . Here , we included the presence of VEGF receptors and co-receptor NRP1 on the luminal endothelial surface based on the evidence that VEGFR2 ( Flk-1 , KDR ) is also present on the luminal endothelial surface [12] . We hypothesize that the distribution of VEGF receptors between the abluminal and luminal surfaces of the endothelial cells ( i . e . , present solely on the abluminal endothelial surface; present solely on the luminal endothelial surface; or present on both surfaces of the endothelial cells ) can impact the VEGF ligand distribution in the tissue and in the blood , as well as the VEGF signaling efficiency . The focus of this paper is to investigate the effects of receptor repartition on endothelial cellular surfaces and emphasize the importance of receptor quantification .
This extension of our previous model [9] is useful for exploring the effects of luminal vs . abluminal distribution of VEGF receptors on the endothelial surfaces . We have shown that such configurations can drastically affect the VEGF profile in the tissue and in the blood . First , we have shown that the removal of clearance in the presence of lymphatics could reverse the free VEGF gradient between the tissue and the blood compartments . Such a situation might correspond to certain pathological conditions , but the simulation is also instructive as a characterization of the VEGF transport system . However , it is important to note that our current model does not explicitly include the convective component of transvascular permeability and such addition could attenuate the predicted gradient reversal . Secondly , at a fixed VEGF secretion rate , the free VEGF in the available interstitial fluid is much higher than that in the plasma . When the free VEGF concentration in the plasma is constant ( ∼1 pM ) , VEGF extravasation and plasma VEGF clearance over time are constant over the range of receptors we studied . We have found that the amount of VEGF disappearing by internalization of luminal receptors to which it binds , the amount of VEGF extravasating and the amount of VEGF removal from lymphatic drainage are all proportional to the luminal receptor density but insensitive to the abluminal receptor density . We have established a mathematical relationship between the amount of VEGF secreted and VEGF disappearing by internalization of abluminal receptors . Thirdly , we can summarize the VEGF transport between the tissue and the blood as shown in Figure 10 . VEGF is secreted in the tissue . Depending on the receptor density on the abluminal and luminal endothelial surfaces , VEGF is mainly either sequestered by the matrix or binds to abluminal receptors . Upon binding , VEGF disappears by internalization of the abluminal receptors it has bound to . Only a small fraction ( free ligands ) enters the blood compartment ( mainly by intravasation rather than lymphatic drainage ) . VEGF then disappears either by internalization of receptors located on the luminal endothelial surface to which they bind or , when the receptor densities are very low , by plasma clearance . This overall transport explains why , regardless of where the receptors are expressed on the endothelial cells ( abluminal vs . luminal surfaces ) , the binding to the receptors occurs more in the tissue than in the plasma ( since a higher concentration of free ligands is available in this compartment – due to secretion – as compared to the free VEGF in the blood ) . However , our simulations have revealed that for high abluminal and low luminal receptor densities , VEGF can bind “preferentially” to VEGFR1 on the abluminal surface and to VEGFR2 on the luminal surface of the endothelial cells . This result requires experimental exploration . In particular , this result shows that quantification of luminal vs . abluminal receptors can be crucial in understanding VEGF signaling in both physiological and pathological conditions . Finally , our simulations reveal that VEGF binds “preferentially” to VEGFR2 compared to VEGFR1 . If VEGFR2 is shown to be pro-angiogenic and VEGFR1 is shown to be anti-angiogenic , then we can conclude that , overall , the signaling is mainly pro-angiogenic regardless of the receptor distribution on the endothelial cells . Since VEGF receptor distribution between the abluminal and luminal endothelial surfaces plays such an important role , it would be interesting to investigate if some pathologies could be explained by decreased receptor expression or internalization . For example , in our previous model , we had shown that an increase in VEGF vascular permeability or secretion could not solely explain the increase of free VEGF concentration in plasma seen in cancer patients [9] . It could be interesting to see if deregulated receptor expression could explain the plasma VEGF increase in cancer ( as compared to healthy subjects ) . The present model suggests , for example , that VEGF could intravasate in high proportion if the amount of VEGF disappearing by internalization of bound receptors decreases , i . e . , if the internalization rate of the receptors or if the receptors expression decreases . The present model also suggests that , since most of VEGF disappears via internalization of bound receptors ( whether on the luminal or abluminal endothelial surface ) , the increase of internalization of receptors could potentially decrease VEGF signal transduction . This could be done either by increasing the internalization rate of the already-existing receptors or by bioengineering cells expressing VEGF receptors which would have the property of having a high binding affinity for VEGF as well as a higher internalization rates than endothelial cells . Decreasing the VEGF signal transduction of endothelial cells could have potential therapeutic applications . For a complex system such as the VEGF receptor-ligand interactions and transport considered , it is necessary to add elements and further increase the degree of complexity step by step in order to understand the effect of each factor . We can outline further steps in refining the model . First , the model has looked at the effect of the receptors in the proportion 1∶1∶1 for VEGFR1∶VEGFR2∶NRP1 . It would also be of interest to see how unequal ratios of receptors can influence the distribution and concentration of VEGF , especially when experimental data on receptor distribution in vivo become available . Secondly , at the moment , the model considers two isoforms of VEGF: VEGF121 and VEGF165 . Other isoforms could be added to the computational model when new quantitative information becomes available . The model could also include neuropilin-2 which could compete for VEGF . Thirdly , the introduction of soluble VEGFR1 ( sFlt-1 ) would also be of interest , especially since recent results have shown that sFlt-1 can serve as an additional means for VEGF to be transported from the plasma into the tissue [7] . In that study , we hypothesized that the anti-angiogenic potential of sVEGFR1 may stem from its dominant-negative heterodimerization with cell surface VEGFRs and predicted that the circulating ( plasma ) level of sVEGFR1 is significantly higher than its interstitial concentrations , which could imply that sVEGFR1 may have a greater modulatory influence on luminal VEGFRs than abluminal VEGFRs [7] , [11] . Platelets have been shown to be significant reservoirs of VEGF in the blood circulation . It would be interesting to include such elements into the model . Again , quantification of luminal receptors would be crucial , especially since platelets have been shown to sequester large amounts of VEGF and release VEGF from α-granules [20] , [21] . Similarly , the body tissue compartment was considered to have the properties of skeletal muscle . It could be important to distinguish between highly vascularized and relatively avascular organs , as well as elements with varying rates of lymphatic drainage . This would require experimental data on VEGF secretion and other tissue characteristics that at present are poorly known . Furthermore , luminal and abluminal receptors may not be equally accessible by VEGF possibly because of endothelial cell polarity: basement membrane on the abluminal side and glycocalyx on the luminal side . A current assumption was the conservation of total ( free and bound ) density of receptors at each time step . In other words , we assumed that the internalization of receptors was equal to the receptor insertion per abluminal or luminal endothelial surface for each time point . Relaxing such assumptions and replacing them by the experimentally-based receptor dynamics would make the model more accurate . In our model , we assumed that the vascular permeability was fixed . In reality , VEGF , also known as VPF ( vascular permeability factor ) , plays an important role in regulating permeability [22] . An addition to the model would be to determine a quantitative relationship between the vascular permeability and the concentration of VEGF and include that relationship in the model . Our study has shown that quantification of luminal vs . abluminal receptors could be very useful to better understand VEGF signaling and the mechanisms underlying VEGF-dependent diseases as well as angiogenesis and will motivate experimental exploration .
The authors thank Elena Rosca , Amina Qutub , Emmanouil Karagiannis , Princess Imoukhuede , Jacob Koskimaki and Prakash Vempati for useful discussions .
|
Angiogenesis is the growth of new blood vessels from pre-existing vasculature that occurs in physiological ( e . g . , exercise ) and pathological contexts ( e . g . , cancer ) . This process is often triggered by a signaling cascade that occurs upon ligand-receptor binding between vascular endothelial growth factor ( VEGF ) and its receptors ( VEGFR1/Flt-1 , VEGFR2/KDR ) . These receptors are expressed by endothelial cells that line the blood vessels . Little is known about the quantitative proportion of abluminal receptors ( facing the tissue ) as compared to those on the luminal surface ( facing the blood ) . We have built a compartment model with molecular details from human tissues to investigate why such experimental data would be of importance . We conclude that the receptor distribution on the endothelial cells can significantly alter the VEGF distribution and the VEGF signaling ( through its binding to the receptors ) and that quantification of luminal vs . abluminal VEGF receptors would shed light on VEGF signaling and VEGF-dependent mechanisms of angiogenesis .
|
[
"Abstract",
"Introduction",
"Discussion",
"Acknowledgments"
] |
[
"computational",
"biology/systems",
"biology"
] |
2009
|
The Presence of VEGF Receptors on the Luminal Surface of Endothelial Cells Affects VEGF Distribution and VEGF Signaling
|
Genome-wide association studies ( GWAS ) have identified hundreds of associated loci across many common diseases . Most risk variants identified by GWAS will merely be tags for as-yet-unknown causal variants . It is therefore possible that identification of the causal variant , by fine mapping , will identify alleles with larger effects on genetic risk than those currently estimated from GWAS replication studies . We show that under plausible assumptions , whilst the majority of the per-allele relative risks ( RR ) estimated from GWAS data will be close to the true risk at the causal variant , some could be considerable underestimates . For example , for an estimated RR in the range 1 . 2–1 . 3 , there is approximately a 38% chance that it exceeds 1 . 4 and a 10% chance that it is over 2 . We show how these probabilities can vary depending on the true effects associated with low-frequency variants and on the minor allele frequency ( MAF ) of the most associated SNP . We investigate the consequences of the underestimation of effect sizes for predictions of an individual's disease risk and interpret our results for the design of fine mapping experiments . Although these effects mean that the amount of heritability explained by known GWAS loci is expected to be larger than current projections , this increase is likely to explain a relatively small amount of the so-called “missing” heritability .
Genome-wide association studies ( GWAS ) have been extremely successful across many diseases in identifying loci harbouring genetic variants that affect disease susceptibility . Virtually all associated variants identified from GWAS to date have relatively small effects: each additional copy of the risk allele typically increases disease risk by 10%–30% ( see for example [1] ) . It has become clear that the variants discovered thus far account for only a small proportion of the genetic basis of each of the diseases , and there has been considerable speculation about where the “missing” heritability might lie [1] . One of several important factors in the success of the GWAS design has been the pattern of linkage disequilibrium in human populations . The strong correlations between nearby SNPs mean that commercially available genotyping chips , which assay 300 , 000–1 , 000 , 000 SNPs , can capture much of the common variation in the human genome , particularly in Caucasian populations [2] . Because genotypes at the causative loci will often be correlated with those at SNPs that are typed on the genotyping chip , it is typically not necessary to assay the true causative variant directly in order to detect a genetic association with disease . While linkage disequilibrium is extremely helpful for GWAS discovery , the downside is that in most reported regions of association , the true causal variant or variants remain unknown . Therefore it is possible that many of the associated SNPs are only surrogates for the true causal variant ( s ) . When it comes to quantifying the genetic effect , the genotype at the reported SNP acts as a noisy measurement of the genotype at the causal variant . This noise can dilute the apparent strength of the effect , and obscure the true relationship between genotype and phenotype . As we progress towards the identification of the causal variants , estimates of effect sizes for associated loci will thus tend to increase . In turn , the proportion of disease susceptibility explained by GWAS loci will also increase . Thus in addition to other plausible sources , such as secondary signals in GWAS loci , rare variants ( <1% frequency ) , copy number polymorphisms , and epigenetic effects , some of the missing heritability is actually contained in loci already identified by GWAS , and is driven by common variation ( >1% frequency ) . In this paper we use an extensive simulation study to investigate , and quantify , this phenomenon . We show that estimates of the size of the genetic effect based on the best SNP from the GWAS genotyping chip can often closely approximate the effect size at the true causal SNP . In some cases the causal SNP has a large effect and is poorly tagged , leading to substantial underestimation of the true effect size . We investigate how much of the “missing” heritability could thus be hidden in reported GWAS loci , under several sets of assumptions about the nature of the effects at true causal SNPs . Our results also inform the design and value of fine mapping experiments in GWAS loci .
To begin , we compare the estimated effect size at the replicated hit SNP with the true effect size at the causal SNP in the simulation . Figure 1 illustrates this comparison for three different values of the true effect size . For each we see a peak of estimates around the true effect size assumed at the causal SNP . But note also that there is often underestimation of the true effect size ( mean estimated effect size 1 . 24 , 1 . 86 and 3 . 32 for true relative risk of 1 . 25 , 2 and 4 respectively ) , and that this underestimation can be substantial when the true effect is large . For example , when the true relative risk is 4 , the estimated effect size was less than two in 12% of simulations of successful GWAS discovery of the effect . In Figure 2 we plot the relative under- ( or over- ) estimation of the effect size ( estimated effect size divided by true effect size ) as a function of the correlation ( as measured by the r2 which is the square of Pearson's correlation coefficient ) between the hit SNP and the true causal variant . The underestimation is seen to be due to imperfect tagging: when the true causal variant is not well tagged by SNPs on the genotyping chip ( the correlation is weak ) , the estimated effect at the hit SNP is often much lower than the true effect . Conversely , when the causal SNP is well tagged by a SNP on the chip , the estimated effects cluster around the true effect size . Note that while underestimation decreases as the correlation between the hit SNP and the causal SNP increases , there remains systematic underestimation even when the hit SNP has r2≈0 . 8 with the causative SNP . For example in one third of simulations when the true effect is two , the estimated effect will be under 1 . 8 . Note also that when the true effect size is large , significant and replicable associations can be detected when the best tag SNP only has r2≈0 . 2 with the causal variant ( Figure 2 , relative risk = 4 ) . Imperfect tagging and an ascertainment effect also explain the feature of the plots whereby the underestimation is much less for smaller true effect sizes . If the true effect is small and the true causal variant is not well-tagged on the genotyping chip , there will not be enough power for the GWAS and subsequent replication to reach significance [5] , with the result that the corresponding simulation will not contribute to the plot . But if the true effect is large there may still be power to see a significant result when the true variant is not well tagged , so the simulation contributes to the plot and shows the underestimation . Put another way , if the true effect is small , it will only be detected in an association study if the causal SNP is well tagged , and in this case the effect size will be estimated reasonably well . This second ascertainment effect explains the lack of underestimation at hit SNPs not strongly correlated to the causal SNP in the left panel of the Figure 2 . Lastly , as low frequency SNPs are less well tagged by other SNPs [6] , the extent of the underestimation also depends on the frequency of the risk allele ( see Figure S1 ) . Interestingly , the effect sizes at rare alleles are underestimated to a great extent , but only when the true effect size is large enough for the tag SNP of a rare allele to be detected and replicated in the simulated GWAS . The results above describe the distribution of estimated effect sizes as a function of known true effect sizes and the frequency of the risk allele . In practice we are actually interested in the reverse question , namely what true effect sizes are plausible in the light of the effect size actually estimated from a GWAS and follow-up study ? We will see that this requires assumptions about the true distribution of effect sizes . Indeed , writing RR for relative risk , and RAF ( risk allele frequency ) for the allele frequency at the risk allele , application of Bayes' theorem gives ( 1 ) where “true” refers to the value at the causal SNP and “observed” refers to the value at the hit SNP . Our simulation study allows us to estimate the first factor on the right hand side of ( 1 ) , and we do so by discretising both the observed and true RR and RAF and creating a matrix of counts based on our simulations over the ENCODE regions . The second factor on the right hand side of ( 1 ) is the assumed joint distribution of true risk allele frequencies and effect sizes , which is of course unknown . We proceed by making two different sets of assumptions about these unknowns . In each case we assume that the distribution of risk allele frequencies is given by the empirical distribution of allele frequencies in the ENCODE regions . In effect this assumes that any SNP variant is , a priori , equally likely to affect disease status . What differs between the sets of assumptions is the assumed effect size of a particular variant . Our first set of assumptions posits that the distribution of effect sizes is the same for all putative causal variants , regardless of their allele frequency , and that effect sizes are close to those observed in GWAS studies . The second set of assumptions explicitly assumes that there might be substantially larger effects at variants with smaller minor allele frequency . These priors are described in detail in the Methods section . Different sets of assumptions about true effect sizes and risk allele frequencies necessarily lead to different conclusions , and it is impossible to study all possibilities . A number of theoretical analyses [7] , [8] , [9] , [10] have argued for a relationship between effect size , disease model , and minor allele frequency ( MAF ) . As there is no consensus on the exact form and extent of the relationship we do not rely on them explicitly here , and instead our approach aims to capture two different perspectives on unknown effect sizes , with the subsequent analyses indicating a range of possibilities . The first perspective is that the range of true effect sizes will be close to those estimated from current GWAS . The second captures the possibility that low-frequency variants may have considerably larger effect sizes . Under either set of assumptions , we can use our simulation study , and Bayes' Theorem ( 1 ) to estimate the conditional distribution of true effect sizes and risk allele frequency ( RAF ) in the light of the observed data at the GWAS hit SNP . Figure 3 illustrates this , showing estimates of the posterior distribution of the true effect size conditional on observing a risk estimate between 1 . 2 and 1 . 3 , for different observed risk allele frequencies , and under the two different prior assumptions on effect size distributions . A common feature of the histograms in Figure 3 is that the mode of the posterior distribution on the true effect size is on , or very closes , to the observed estimate . That is , current estimates from GWAS studies of effect sizes from a common SNP , in the range 1 . 2–1 . 3 are most likely to be very close to truth . As expected , estimated effects within this range are more likely to be 1 . 3 than 1 . 2 , because larger effects are more likely to generate a signal of association strong enough to pass the p-value thresholds commonly implemented in GWAS . This explains the left hand tail of the distributions represented in Figure 3 . Figure 3 also shows that there is some probability that the effect size at the causal variant is greater than estimated from the most associated SNP . Interestingly , the observed risk allele frequency impacts our posterior belief about the true effect size , under either set of prior assumptions , with underestimation be more marked when the risk allele at the hit SNP is rarer . Under the conservative prior , when the risk allele at the hit SNP has less than 20% frequency in the control population , the probability that the relative risk is above 1 . 325 is 55% , compared to 35% when the risk allele frequency is between 20–50% . The corresponding numbers for the MAF-dependent prior are 77% and 49% . There are several different phenomena at work here . If the hit SNP is the causal SNP then , assuming that the association is strong enough to be detected and replicated in the GWAS , there is no systematic under estimation ( and very little over estimation as we assume the effect size is estimated from the replication sample ) . However , conditional on the hit SNP not being causal , the distribution of LD with true causal SNP , and therefore the propensity for under estimation , depends on its allele frequency . The posterior distribution on the true effect size given the observed frequency and effect of the hit SNP can be viewed as a mixture of these two scenarios , weighted by their conditional probability . Rarer SNPs are less likely to be tagged well by single markers , and as noted above , poor tagging leads to underestimation of effect sizes . In contrast , for a common SNP , the associated allele is more likely to be well correlated with the causal allele , so there is relatively less under estimation . Under the MAF-dependent prior , when the associated allele is low-frequency the causative allele will tend to be low-frequency as well , and so potentially of larger effect . In the scenario where we believe in larger effects at rare causal alleles and have observed a SNP with low RAF with estimated relative risk between 1 . 2 and 1 . 3 there is a 24% chance that the source of the signal is a variant which actually doubles or more than doubles risk with each copy of the risk allele . Our observations are similar when the observed risk allele is the most common allele in the population ( RAF>50% ) and therefore the minor allele is protective ( Figure S2 ) . Qualitatively , the same conclusions also apply when the estimated effect size at the hit SNP is weaker , for example in the range 1 . 05 to 1 . 2 ( Figure S3 ) . One consequence of the potential underestimation of effect sizes from GWAS findings is that as we move to better identification of the actual causal variants , through fine mapping and/or functional studies of associated regions , our estimates of their effect sizes might well increase . Assuming a multiplicative model of risk across loci , these small expected changes could combine to increase the relative risk of disease in those individuals with highest genetic risk of disease . To investigate this , we simulated genotypes at known associated loci in a population of individuals ( assuming Hardy Weinberg equilibrium and no linkage disequilibrium across loci ) for each of breast cancer , type 2 diabetes and Crohn's disease , based on reported risk allele frequencies [11] , [12] , [13] ( see Tables S3 , S4 , S5 for a list of loci ) . First treating the causal loci and relative risks for each disease as given by current GWAS estimates , we measured the average risk of individuals in the top x% , by risk , of the population ( for differing values of x ) and compared this to the mean risk in the population . We then repeated this simulation , allowing for the uncertainty in the estimation of true effect sizes by averaging over the uncertainty in both the RAF and effect size of the causal variant on the basis of the posterior distributions of these , given the GWAS findings , under the two priors described above . We assumed that risks combined multiplicatively across loci . For NOD2 and IL23R in Crohn's disease where the causal variant is thought to be known , here and below , we used the effect sizes for the known variant , and did not average over uncertainty in these . Because all three diseases have been extensively studied , we approximated the GWAS discovery process as corresponding to a GWAS discovery sample of 5000 cases and 5000 controls , and a replication sample of 10 , 000 cases and controls . The actual discovery process for each of the diseases is complicated , often involving meta-analysis and/or multistage discovery , and not straightforward to model accurately , but the approach we use should capture the fact that GWAS-discovery were ascertained through study of large numbers of samples . The results of the three simulations are given in Table 1 . The unadjusted simulations give estimates of how much more at risk individuals with the greatest genetic propensity to disease are , based only on GWAS loci , relative to the average person in the population . As expected , the fold change in risk of individuals carrying a large fraction of risk variants is dependent on the number and magnitude of known loci . For example , individuals in the top 0 . 1% of risk for Crohn's disease are 20 times more likely than the average person to develop the condition , whereas for breast cancer , where the number of common loci and associated relative risks is typically smaller , the equivalent number is just over two-fold . The second and third simulations attempt to average over the possible outcomes of our future efforts to map causal mutations , to reveal the likely gains in our ability to stratify individuals on the basis of risk . These use the methodology above , under both prior distributions , to average over the posterior distribution of the allele frequency and effect size at the causal SNPs underlying reported GWAS loci for the three diseases . These adjusted estimates are also shown in Table 1 . Across diseases we see that there is a significant increase in the risk associated with carrying multiple risk variants . In particular we see that the biggest differences in risk are for those individuals in the extreme tail . It is these individuals who carry the stronger , likely rarer , risk alleles which are currently insufficiently characterised by the most significant signal of association in some regions identified to be important in disease . For example , the risk of an individual in the top 0 . 1% of the population for genetic risk typed at the causal loci underlying currently known GWAS loci will likely be increased by a factor of 3–6 . 5 , 5–12 , or 25–50 , compared to an average individual , for breast cancer , type 2 diabetes and Crohn's disease . These are notably greater increases in risk than current prediction based in the hit SNPs from GWAS loci which would be 2 . 4 , 3 . 5 and 20 respectively . We have shown above that as we move to identification of the true causal variants underlying GWAS associations , through fine mapping and functional studies , their effect sizes will tend to increase , in a minority of cases substantially , compared to current estimates from GWAS . This will , in turn , increase the amount of heritability explained by these diseases . We can use the approach developed here to try to quantify this effect . We investigated this question in the context of the three diseases just described , namely breast cancer , type 2 diabetes , and Crohn's disease . For each disease we took the set of hit SNPs from published associated loci [11] , [12] , [13] ( see Tables S3 , S4 , S5 ) , and for our two prior distributions on effect sizes we estimated the posterior distribution of both the effect size and the allele frequency for the causal SNP at each locus , as described in the previous section . One commonly used measure of heritability is sibling recurrence risk ratio , often denoted by λS: the relative increase in risk to an individual if their sibling has the disease compared to the baseline risk in the population as a whole [14] . Assuming , as is usual for heritability calculations [15] , that there is no interaction between loci , λS can be calculated as a function of the risk allele frequency and effect size for each causal variant . In order to allow for the uncertainty in the allele frequency and likely underestimation of the effect size at the causal variants underlying GWAS associations , we averaged this expression over the posterior distribution of these quantities , given the GWAS findings ( see Methods for details ) . The results are shown in Figure 4 . For each disease they show that the heritability due to already identified GWAS loci will be higher than current estimates , under either set of assumptions about true effect sizes , but particularly under the MAF-dependent prior . Whereas at the time of writing the current estimates of the contribution to λS from GWAS loci are 1 . 03 , 1 . 08 , and 1 . 49 for breast cancer , type 2 diabetes , and Crohn's disease , these may well be 1 . 06 , 1 . 14 , and 1 . 61 ( mean under the conservative prior ) and they could plausibly be as high as 1 . 21 , 1 . 39 and 2 . 46 ( mean under the MAF-dependent prior ) . Whilst some of the “missing” heritability is thus disguised rather than missing , we note that this effect is unlikely to account for the extent of the gap between estimates of sibling relative risk ( 2 , 1 . 8 , and 10 , respectively , from family studies [16] , [17] , [18] ) and those explained by currently known loci . We return below to a discussion of the discrepancy .
The correlation between alleles along the human genome has allowed GWAS to look for regions associated with disease without having to either genotype all known genetic variation or guess a priori which regions of the genome may be important . Although this approach has been a significant success , there is a predictable downside of using a subset of variation to tag , or predict , untyped diversity: for the vast majority of the SNPs identified as mediating disease risk , we are left uncertain as to whether they are causally involved in the pathway from genotype to phenotype , or , much more plausibly , are just a surrogate for the causal variation . GWAS associations will thus typically relate to a noisy measurement of the causal variant . One consequence of this is that the size of the genetic effect associated with GWAS loci may be underestimated . We quantified this through an extensive simulation study designed to mimic patterns of linkage disequilibrium in European Caucasian populations . We draw two broad conclusions from these analyses . Firstly , a significant proportion of estimated relative risks will be biased downwards because the hit SNP is a powerful , but imperfect , tag for the true causal variation . In most cases this effect will be relatively minor , but in some instances , the best associated SNP may actually be a poor predictor of a , putatively rarer , SNP with a much larger effect , in which case the effect size estimated from the GWAS finding will substantially underestimate the true effect size . The exact proportion of reported associations which fall into these two categories depends on properties of the design of the study from which the SNP was identified , and on one's belief about how likely low frequency ( >1% ) variants of large effect are to cause common diseases . The statistical power afforded by any particular association strategy sets a lower limit on the size of effect that can be under-estimated because an imperfect tag of an allele with a small effect size will simply fail to achieve genome-wide significance . Other properties of GWAS strategy , such as sample ancestry and the number of markers typed , also change our interpretation of observed effect sizes because they influence the distribution of linkage disequilibrium between putative hit SNPs and causal variants . Our findings show that at any particular locus , especially if the associated SNP has a low MAF , the true effect could be quite large . But we would not expect this to be widespread . Were many true effects this large it would be extremely surprising for so few of them to have been observed: although any one such causal SNP may not be well tagged on the genotyping chips used for GWAS , some of them will happen to be at least moderately well tagged , and their detection would lead to much larger estimates than have been seen from current studies . In the context of this study these early observations suggest that , of the two prior distributions we investigated , it is the conservative prior that may better reflect the true distribution of effect sizes attributed to low and common frequency variants . One way of viewing the posterior distribution on the true effects shown in Figure 3 is as a probability distribution on the outcome of efforts to fine map current regions of association . In this light , our results inform questions of the design and value of fine mapping experiments . First , simulations similar to those described above ( assuming causal variation to be distributed like SNPs in ENCODE regions ) suggest that less than 8% of the time will the hit SNP actually be the causal SNP . We note that there may be more reward in terms of gains in predictive ability and increases in effect size from fine mapping SNPs with lower minor allele frequency because they are , on average , more likely to be in poor LD with an unobserved causal variant . On the other hand , our simulations show that although they are unlikely to be causal , most common hit SNPs are likely to be very good surrogates markers for their causal variant . Indeed , in 25% of cases , the hit SNP will be a near-perfect surrogate ( ie r2>0 . 99 ) for the causal variant . Should this be the case , further genotyping will not reveal other SNPs with stronger associations , unless sample sizes are extremely large . Here we have quantified the increased spread of genetic risk with genotypes just at known loci , and only considering a multiplicative disease model . But even in this restricted setting , there will be substantial differences in risk between high- and low-risk groups based on these genetic factors . For example the propensity of individuals in the top 0 . 1% of the population distribution of genetic risk of type 2 diabetes will be increased by a factor of 5–10 , compared to the average . For breast cancer , in the analogous top-risk group this risk will be increased by a factor of 3–5 ( on the basis of common variation ) . Importantly , with the growth of GWAS findings , both in terms of numbers of diseases and numbers of loci for particular diseases , more and more of the population will be in this most at risk category for at least one disease: assuming 100 independent diseases , nearly 10% of the population will be in the top 0 . 1% of risk of at least one disease . Knowing which individuals these are and what diseases they are most at risk of is therefore potentially useful information , both to the individual and at the population level . The issues involved in utilising such information in screening programmes ( discussed for example in [13] ) are complicated , but our results strengthen the arguments for consideration of this possibility . We have shown that some of the “missing” heritability for common disease actually resides in known GWAS loci and have estimated this deficit for three particular diseases . While rather more heritability is likely to be explained by known GWAS loci than has been reported , this effect alone falls well short of explaining all the missing heritability . Note , however , that there are other reasons why existing loci may explain more heritability than currently thought . Current calculations ( by others , and above ) focus on a single causal variant in each associated region: more variants within regions will explain more heritability . They also ignore possible non-multiplicative disease effects , and also ignore interactions between variants at different loci . Power to detect either is low [19] , so it is misleading to put much weight on the failure of existing designs to find such effects . As others have noted [20] , parts of the missing heritability could be due to multiple rare variants of large effect , associations with other forms of genetic variation such as copy number polymorphisms , and epigenetic effects . Indeed it would be surprising if each did not play some role . Another possibility is that estimates of the “genetic” component of disease susceptibility , from epidemiological studies , confound shared environment with shared DNA , and so inflate heritability estimates [21] , [22] .
In order to model the signal of association generated by disease-causing mutations , we chose to simulate data exploiting empirical surveys of human diversity . For this purpose we used data from the 10 ENCODE regions [23] within the CEU analysis panel of HapMap II [5] , which have undergone SNP ascertainment by resequencing 48 individuals of diverse ancestry . These regions therefore show a fuller spectrum of SNPs than are represented in the HapMap data at large , and haplotypes are expected to be accurate due to the trio design of the CEU HapMap panel [24] . The regions over which we simulate data are centred on each of the 10 ENCODE regions ( listed in Table S1 ) and include 500kb of flanking HapMap variation at the boundaries of each region . As the typical sample size of most GWAS is much larger than the number of CEU HapMap individuals , we simulated 100 , 000 chromosomes using the HAPGEN software package . These 100 , 000 haplotypes we call the reference panel . GWAS case and control samples were then subsampled from the reference panel , as described below . HAPGEN uses a population genetic model that incorporates the processes of mutation and fine-scale recombination to generate individuals from an existing set of known haplotypes . We ran HAPGEN with an effective population size of 11418 ( as recommended for the CEU population ) , a population scaled mutation rate of 1 per SNP , a population scaled recombination rate from estimates described in [25] , with the known set of haplotypes taken from the CEU analysis panel of HapMap II as described above ( see http://www . stats . ox . ac . uk/~marchini/software/gwas/hapgen . html ) . For SNPs greater than 1% in frequency in the ENCODE regions we performed two hypothetical GWAS by letting each of the two alleles be causal in turn . We denote the causal allele by A and the protective allele by a . To generate the control sample we sampled the required number of haplotypes , without replacement , from the reference panel and combined these in pairs to form diploid individuals . This mimics the common use of population controls , rather than controls explicitly chosen for not having the disease under study . For the case sample , we sampled pairs of haplotypes from the reference panel according to the genotype frequencies at the causal SNP dictated by the assumed disease model: If δ is the risk of the AA genotype , and α is the risk of the Aa genotype , both relative to the aa genotype , then we sample case individuals ( without replacement ) on the basis of their genotypes at the SNP assumed to be causal with success probabilities proportional to: ( 2 ) where f is the frequency of the risk allele A in the reference panel . Throughout , for definiteness , we adopted a multiplicative model for disease risk ( additive on the log scale ) defined by δ = α2 . We refer to α as the relative risk ( RR ) or effect size associated with the causal variant . To approximate a GWAS , we thinned the generated data set to include only those SNPs present on the Affymetrix 500K array that had a minor allele frequency in sampled controls of greater than 1% . This set may or may not include the assumed causal SNP . For analyses involving only simulated data , we sampled 2 , 000 cases and 2 , 000 controls from the reference panel to emulate a typical large GWAS . For the subsequent analyses of heritability and individual risk profiling for type 2 diabetes , breast cancer and Crohn's disease that studied particular reported associations , we simulated 5 , 000 cases and 5 , 000 controls to obtain results more comparable to the size of study from which the associations were ascertained . We simulated under a range of relative risks at 24 grid points from 1 . 05 to 6 . In attempting to simulate the signal of disease at rare alleles ( 1% to 5% ) in a GWAS of 5000 cases and controls there were a small number of simulations in which there were insufficient haplotypes in our reference panel to generate the required number of genotypes at the causal SNP for large effect sizes . These simulations were discarded , but as the numbers were small ( 3% when the RR = 4 and 11% when RR = 6 ) we do not believe this greatly affects the results presented below . Following common practice , for each simulated case control sample , we tested for association between genotype and case control status using the Cochran Armitage trend test [26] at each SNP with frequency greater than 1% in the simulated panel of chromosomes . We calculated the p-value of this test statistic which is distributed with 1 degree of freedom under the null hypothesis of no association . If any test across the region obtained a p-value<10−6 the location of the most significant SNP ( termed the hit SNP ) was recorded and we simulated this SNP in an independent replication sample . We simulated the replication experiment in three stages . First we simulated the frequency of the causal allele in cases and controls in the replication population . We then simulated the frequency of the hit SNP conditional on the frequency of the causal allele . Finally , we simulated the genotype counts for a sample of cases and controls in this replication population . We motivated sampling of the frequency of the causal allele in controls in the replication population by thinking of the replication sample as an additional sample from the same population as the original GWAS sample . ( Other assumptions are possible here , but seem unlikely to affect the main conclusions . ) Specifically , we placed a uniform prior distribution on the unobserved population frequency and sampled a value , f ′ , from the posterior distribution of this frequency given the data in the reference panel . ( Given the large size of the reference panel , the frequency in the replication sample will be very close to that in the reference panel . ) Conditional on f ′ , the population replication frequency in cases was calculated from equation ( 2 ) . To obtain the replication population frequencies at the hit SNP we estimated the conditional distribution in the reference sample of alleles at the hit SNP in each of cases and controls , given those at the causal SNP , and used these for the replication sample . This corresponds to assuming that the LD between the causal and hit SNP in the replication sample will be the same as that in the reference sample . Finally , conditional on the population replication frequencies in cases and controls , we take multinomial samples of the required size to mimic the replication case and control samples . A test of association using the trend test was performed at the hit SNP on the simulated replication samples and deemed a significant replication if the p-value was less than 10−2 . We estimate the effect size , or relative risk , α , at the hit SNP by maximum likelihood under the model described above by equation ( 2 ) . For studies with population controls this can be achieved in practice by fitting a logistic regression model for case status [27] . We implement two different sets of prior assumption on the effect size and its relationship with minor allele frequency . Our first set of assumptions is that if α is the effect size at a causal variant , then log ( α ) is normally distributed with mean 0 and standard deviation 0 . 2 , independent of RAF . We refer to this as the conservative prior , since it places little weight on relative risks greater than 1 . 5 . To get a sense for this distribution , it assumes that 81% of true effect sizes are less than 1 . 3 with 96% less than 1 . 5 , and 99 . 9% less than 2 . A further discussion of the choice of prior on effect sizes can be found in [19] and [28] . Our second set of assumptions , which we call the MAF-dependent prior , again assumes a normal distribution for log ( α ) with mean 0 , but here the standard deviation , σ , is allowed to depend on the RAF . The dependence of the distribution of the effect size on allele frequency has no theoretical justification , but is chosen on pragmatic grounds to give a gradual increase in the average effect size as the alleles at causal SNP become rarer in the population . It is implemented by increasing σ by a weight defined by an exponential density with parameters chosen such that , when the RAF is near 0 . 5 ( a common SNP ) , this prior is approximately the same as the conservative prior , with σ = 0 . 2 . As the RAF approaches 0 or 1 ( corresponding to rarer SNPs ) , then considerably more weight is put on larger RRs . See Figure S4 and Table S2 for details . For example , when the MAF is less than 5% the second prior gives an approximately 45% chance that the risk associated with each copy of the causal allele is larger than 2 . 5 . Note that we used an empirical prior on the frequency of the risk allele ( Figure S5 ) by choosing each allele , at each SNP , with in the ENCODE region to be causal in turn . A commonly used measure of heritability is based on considering the risk of disease to an individual conditional on them having an affected ( ) sibling relative to the unconditional probability ( which is just the prevalence of the disease ) :We can calculate the above , using assuming that , and by summing over the genotypes of the siblings and of the mother and father ( see [15] ) :If we divide through by the square of the risk associated with most protective genotype ( which we can define to be ) then we can write the above in terms of the per allele relative risk , and assume the genotype probabilities follow Hardy-Weinberg equilibrium with risk allele frequency as above:By making the further assumption that loci are independent an estimate of the heritability explained by a set of hit SNPs can be obtained by multiplying together the λS values calculated at each individual locus . We calculated sibling relative risk in this manner using estimates of RR and RAF of replicated loci from studies of Type 2 diabetes , Crohn's disease and breast cancer ( see Tables S3 , S4 , S5 ) . We then simulated 100 , 000 times from the posterior of true RR and RAF of each locus conditional upon the reported RR and RAF , using the simulation approach and the two different priors as described in the paper . For each set of simulations , for each disease , we recalculated λS at each locus and multiplied over loci , giving a sample from the posterior distribution of sibling risk that could be explained by the current set of report loci if the causal loci where typed directly .
|
Genome-wide association studies ( GWAS ) exploit the correlation in genetic diversity along chromosomes in order to detect effects on disease risk without having to type causal loci directly . The inevitable downside of this approach is that , when the correlation between the marker and the causal variant is imperfect , the risk associated with carrying the predisposing allele is diluted and its effect is underestimated . Using simulations , where we know the true risk at the causal locus , we quantify the extent of this underestimation . We show that , for loci which have a modest effect on disease risk and are common in the population , the risk estimated from the most associated SNP is very close to the truth approximately two thirds of the time . Although the extent of the underestimation depends on assumptions about the frequency and strength of the risk allele , we predict that fine mapping of GWAS loci will , in rare cases , identify causal variants with considerably higher risk . Using three common diseases as examples , we investigate the expected cumulative effects of underestimation at multiple loci on our ability to stratify individuals by disease risk and to explain disease heritability .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"genetics",
"and",
"genomics/disease",
"models",
"genetics",
"and",
"genomics/genetics",
"of",
"disease",
"genetics",
"and",
"genomics/complex",
"traits"
] |
2011
|
Quantifying the Underestimation of Relative Risks from Genome-Wide Association Studies
|
Epidemiological modelling has a vital role to play in policy planning and prediction for the control of vectors , and hence the subsequent control of vector-borne diseases . To decide between competing policies requires models that can generate accurate predictions , which in turn requires accurate knowledge of vector natural histories . Here we highlight the importance of the distribution of times between life-history events , using short-lived midge species as an example . In particular we focus on the distribution of the extrinsic incubation period ( EIP ) which determines the time between infection and becoming infectious , and the distribution of the length of the gonotrophic cycle which determines the time between successful bites . We show how different assumptions for these periods can radically change the basic reproductive ratio ( R0 ) of an infection and additionally the impact of vector control on the infection . These findings highlight the need for detailed entomological data , based on laboratory experiments and field data , to correctly construct the next-generation of policy-informing models .
The language of probabilities and chance entered mathematical epidemiology , then dubbed pathometry , almost from its foundations [1] . However , the common use , and huge success , of deterministic models for disease dynamics ( ODEs ) have made certain probabilistic assumptions very popular in the literature; for example the implicit assumption that the time between various epidemiologically important events is exponentially distributed , which follows from assuming constant per capita rates of change . For vector-borne diseases ( VBDs ) the ODE-based approach dominates , originating with Ross’ modelling of malaria [2] . In the 1940s and ‘50s Macdonald and Garrett-Jones , amongst others , made substantial contributions to the mathematical theory of VBDs [3 , 4] , including the fundamental appreciation that only a fraction of inoculated vectors will survive the extrinsic incubation period ( sometimes called sporogony ) of a disease to become actively infectious . The probability of surviving the extrinsic incubation period ( EIP ) has been most conveniently expressed by assuming that the EIP duration is fixed [3 , 4] . The epidemiologically important factors in a vector’s life history are its life duration , how many times it bites in its life and how many of those bites are infectious . Variation between vector life histories can be modelled as probabilistic; the difference in outcomes between vectors are modelled as being due to identically distributed chance factors rather than intrinsic variation in vector fitness . This requires estimation of the underlying random distributions governing vector life histories . The popularity in the modelling literature of implicitly assuming that relevant distributions are either exponential or fixed length is often due to the popularity of ODE models and mathematical convenience rather than biological motivation . Ross-Macdonald theory has remained very popular amongst mathematical modellers of VBDs . A century after Ross’ original mathematical analysis , reviews of the modelling literature reveal that modern analysis still typically uses the majority of the core set of assumptions familiar to Ross , Macdonald and Garrett-Jones [5] . This has led to speculation that the entire field could have become ‘canalised’ implying that there exist equally fruitful avenues of research that are being under-explored due to the popularity of the Ross-Macdonald approach [6] . However , Ross-Macdonald theory is essentially an attempt to think clearly and quantitatively about the vector life events that must occur in the host-to-vector-to-host transmission cycle of VBDs . In particular their theory allows the integration of important entomological information [the vector biting rate ( α ) , vector mortality rate ( μ ) and the vector population density relative to hosts ( M ) ] with epidemiological data [the vector competence for the given pathogen ( V ) and the extrinsic incubation period , or inverse incubation rate , ( σ−1 ) ] into epidemic metrics for VBDs that give a prediction for transmission intensity . A popular metric is the vectorial capacity ( C ) , first proposed by Garrett-Jones [4] , which measures the expected number of future infected hosts due to a single infectious host per day [7] . If we assume the host has mean infectious duration 1/γ , which includes host mortality and possible excess mortality due to disease , and additionally assume that the host is certain to survive its incubation period , then combining vectorial capacity with the mean duration of host infectiousness gives the classical reproductive ratio for the VBD as R 0 c = C / γ = M V α 2 γ μ exp ( - μ / σ ) . ( 1 ) We use the notation R 0 c for the classical Macdonald reproductive ratio for a VBD . However , in relationship to other common definitions of R0 ( e . g . [8] ) it could equally be described as R 0 2 as the square of the geometric mean of two infectious generations: vectors and hosts . R 0 c is a threshold quantity for the epidemic; that is if control efforts can induce a situation where R 0 c ≤ 1 for a sustained period the VBD must be eradicated [8] . On the other hand , recent analysis [9 , 10] has suggested that the reproductive ratio for the midge-borne viral disease bluetongue virus ( BTV ) spreading between a single host population and a single midge species can be expressed as , R 0 = M V α 2 γ μ σ σ + μ . ( 2 ) The discrepancy between the two formula is due to the interaction between the lifetime of the vector , modelled as an exponential distribution , and the assumed distribution of the EIP . R 0 c ( eq 1 ) is derived from assuming a constant ( fixed ) duration EIP , whereas R0 ( eq 2 ) is generated by assuming an exponential distribution as epitomised by ODE models . Despite both models assuming the same average EIP duration ( 1/σ ) and vector life expectancy ( 1/μ ) , they predict different probabilities of an inoculated ( female ) vector surviving her EIP and therefore becoming actively infectious . When the incubation rate is much faster than the mortality rate ( σ ≫ μ ) the two R0 predictions become identical because the vector is very likely to survive its EIP and probabilistic details become irrelevant . This will not be true for vectors ( such as midges and sandflies ) that are short lived compared to their typical EIP duration . The classic assumption that the lifetime of a vector is exponentially distributed is equivalent to assuming that it has a constant hazard rate of death . In this we follow Macdonald [11] and argue that due to predation , and other persistent environmental risks to the vector , exponentially distributed lifetimes are a reasonable model for many vectors . Constant hazard is probably a more accurate modelling choice for smaller arthropods ( with less complex life-histories ) such as biting midges or sandflies than compared to their larger cousins such as mosquitos and tsetse flies . Certain populations of the biting midge species Culicoides sonorensis have been found to have a lifetime distribution that closely matches exponential , even in a laboratory setting where predation pressures and other environmental risks are absent [12] . In addition , field estimates of mortality based on the proportion of parous vectors caught implicitly assume exponential lifetimes ( e . g . [13] ) . For larger arthropods there is stronger evidence of senescence [14] , however laboratory studies of this facet represent an upper estimate on the survivorship of arthropods in the natural setting . In contrast , we would question the appropriateness of the exponential distribution as a good model for other biological durations significant to the epidemiology of VBDs . Vectors cannot become immediately infectious after successful inoculation with a VBD due to the time required for the pathogen to escape the mid-gut of the vector and spread to its salivary glands from where it can be transmitted to susceptible hosts; it is this biological process that defines the EIP for the vector . It does not seem plausible that this duration should be identical for each vector , as implied by Eq ( 1 ) but nor does it seem plausible that the EIP should be modelled as ceasing at a constant rate after inoculation , as implied by Eq ( 2 ) . Moreover , vectors such as female biting midges and mosquitos generally only need to successfully feed once in their gonotrophic cycle in order to provide protein for their egg yolks as well as for sustenance . Even if there is a local abundance of suitable hosts and breeding sites , implying that the vector spends only a short period in seeking behaviour , the biting activity of the vector is still limited by the duration of oogenesis ( egg production ) and oviposition ( egg laying ) . Since oogenesis is a complex multi-stage biological process , rather than a persistent ‘risk’ , it is unlikely that the duration between bites by a female midge is exponentially distributed . We introduce a flexible approach to modelling vector life histories as random events during an individual vector’s life . The random waiting time between bites on hosts is drawn from a general distribution , that does not have to be the standard exponential . The generalisation of the vector biting process to non-exponentially distributed gonotrophic cycle duration defines a renewal process , a standard theoretical model for processes of discrete events [15] . The EIP duration is also allowed to be chosen from an arbitrary positive distribution; the probability of a vector surviving its EIP duration can then be expressed analytically . Combining these two insights we develop a numerical procedure for calculating a generalised reproductive ratio ( R 0 g ) , and also an analytic approximation ( R ˜ 0 g ) based on the renewal process for biting rate of longer-lived vectors . Given the uncertainty in the forms of these distributions , it is convenient to introduce a dispersion scale , which interpolates between the fixed-duration , dispersion-zero , gonotrophic cycle assumption Eq ( 1 ) and the exponentially-distributed , dispersion-one , gonotrophic cycle assumption Eq ( 2 ) . This approach does not capture all possible distributional choices , but does include the two most popular [eqs ( 1 ) and ( 2 ) ] . We demonstrate that for short-lived vectors these changes have a significant effect on both predicted intensity of VBD transmission and the efficacy of control measures . Culicoides genus biting midges spreading the multi-serotype orbivirus bluetongue virus ( BTV ) provide an ideal case study for the theory developed , due to the plethora of entomological and virological data available on the midge-BTV disease complex . BTV causes the economically important bluetongue disease amongst ruminants , both wild animals and commercial livestock , and is in particular associated with severe disease and increased mortality amongst sheep . BTV circulates persistently in Africa , North America , Australia , the Middle east , China , the Indian subcontinent and southern Europe wherever culicoides midge species are present . Moreover , the virus has demonstrated epizootic invasion capability into northern Europe ( north of 50°N ) [16] . In 2006 BTV invaded northern European herds of commercial ruminant livestock for the first time in record , and demonstrated the ability to overwinter before being controlled by mass vaccination campaigns of over 100 million animals in the subsequent years [17] . The generalised predictions from this paper show that the effort required to achieve eradication by host vaccination or insecticidal spraying can be substantially greater or lower than those offered by classical Ross-Macdonald theory , depending on the assumed underlying distributions . However , we show that whenever R 0 g agrees with R 0 c , the classical approach overestimates the efficacy of reducing vector life expectancy with adulticidal spraying . As experimental knowledge about midge gonotrophic cycles and EIP increases and becomes more detailed [18 , 19] there should be an concurrent effort from modellers of VBDs to integrate more of this detailed entomology into predictions of disease risk . The theory presented in this paper allows the full distributional information on the length of gonotrophic cycle and EIP to be integrated into a prediction of transmission intensity: this represents a significant development from the classical formulation for R0 [eq ( 1 ) ] where only information about averages can be used to estimated disease risk .
The adult lives of the vectors are modelled as a series of independent and identically distributed ( i . i . d ) random waiting periods—the gonotrophic cycles—which we label Gn ( n ≥ 0 ) . These define the time between each successful bite until death , and death is assumed to occur at constant rate μ ( Fig 1 ) . This means that the lifetime biting process of the vector follows a renewal process [15] stopped at the end of the vector’s life . The starting point of the gonotrophic cycles for each vector is taken to be its emergence as an adult . If a vector is inoculated with the pathogen , then only bites after a random duration E ( the EIP ) are potentially infectious . The product of the probabilities of transmission from infectious host to vector and from infectious vector to host ( the vector competence ) is given as V . The classical assumption is that the Gn are exponentially distributed ( G ∼ exp ( α ) ) and E = 1/σ is constant . Here we drop these assumptions and generate a generalised reproductive ratio ( R 0 g ) for a single midge species spreading a pathogen amongst a single host species ( Fig 1 ) . Assuming that the average host infectious duration is 1/γ , R 0 g expressed in explicitly probabilistic terms: R 0 g = V γ × E [ rate of vector population bites per host ] × E [ # bites after EIP by inoculated vector ] . ( 3 ) The factorisation in eq ( 3 ) is appropriate because the future biting of a vector after its inoculation can be taken as independent of its past life-history . The probability of a vector surviving its EIP duration and the average biting rate over the vector population can be solved in terms of the moment generating functions ( MGFs ) of the random life-cycle distributions for the vector . The MGF for a random variable X is defined as , ϕ X ( θ ) = E [ exp { - θ X } ] . ( 4 ) From an entomological modelling perspective the MGF evaluated at θ > 0 is the probability of a vector surviving a random period defined by X whilst undergoing constant mortality rate θ . The two MGFs we will use in our analysis are ϕE , the MGF of the EIP duration , and ϕG , the MGF of the gonotrophic cycle duration . Therefore the probability of a vector surviving its EIP , P E = ϕ E ( μ ) . ( 5 ) The classical assumption that the EIP is fixed at σ−1 for each inoculated vector in fact leads to the smallest estimate of PE over any choice of distribution for the EIP with average duration σ−1 [eq ( 15 ) ]; that is the classical estimate is almost certainly an over-estimate for the proportion of inoculated vectors that die before becoming actively infectious . The average number of bites , or equivalently survived gonotrophic cycles , in the lifespan of a vector is , E [ number of lifetime bites ] = ϕ G ( μ ) 1 - ϕ G ( μ ) . ( 6 ) A traditional modelling assumption is that each vector bites according to a Poisson process at rate α ( each gonotrophic cycle is exponentially distributed with a mean cycle duration of 1/α ) giving a population biting rate per host of Mα where M is the vector-to-host population ratio . A convenient feature of modelling vector bites as a Poisson process is that the chance of the vector biting at any moment in time is completely independent of the vector age , a , since emerging as an adult . This is not true in general; it is biologically plausible that having bitten successfully a vector is less likely to bite shortly afterwards , even if the average time between bites remains 1/α . This could be experimentally confirmed by observing that gonotrophic cycles are under-dispersed compared to an exponential distribution . As a consequence vectors of different ages will have different chances of biting: vector age dependence in the model emerges from considering non-exponentially distributed gonotrophic cycles without any assumption on vector senescence . We will demonstrate that for a general choice of gonotrophic cycle distribution the equilibrium biting rate of the vector population upon hosts is the expected total number of bites made in per vector lifetime divided by the vector life expectancy . This has a concise mathematical expression: E [ rate of vector population bites per host ] = M μ ϕ G ( μ ) 1 - ϕ G ( μ ) . ( 7 ) Eq ( 7 ) is consistent with the classical prediction of biting rating from the vector population in that if the biting process is Poisson ( waiting times G are exponentially distributed ) then , ϕ G ( μ ) = α α + μ ⇒ M μ ϕ G ( μ ) 1 - ϕ G ( μ ) = M α . ( 8 ) Although eq ( 7 ) holds in general , eq ( 8 ) is specific to the Poisson assumption . A second quantity that is key to calculating the basic reproductive ratio is the expected number of infectious vector bites . Infectious bites are those that occur after the latent EIP duration—we denote this quantity BI ( see eq ( 20 ) below ) . In order to make at least one infectious bite an inoculated vector must survive its EIP and any remaining gonotrophic period after its EIP finishes . This implies three key considerations: Because a vector’s inoculating bite also initiates its EIP it is possibly to calculate BI exactly using an integral formula [eq ( 20 ) ] . We can also approximate BI explicitly , using the asymptotic distribution of remaining time until next bite , E [ # bites after EIP by inoculated vector ] = B I ≈ ϕ E ( μ ) α μ . ( 9 ) The approximation is exact if the gonotrophic cycle is exponentially distributed . Putting Eqs ( 7 ) and ( 9 ) together gives the generalised reproductive ratio: R 0 g = M V γ μ ϕ G ( μ ) 1 - ϕ G ( μ ) B I , ( 10 ) and the approximate generalised reproductive ratio ( R ˜ 0 g ) : R ˜ 0 g = M V α γ ϕ E ( μ ) ϕ G ( μ ) 1 - ϕ G ( μ ) . ( 11 ) Eq ( 11 ) explains the discrepancy in the introduction . If the EIP duration is fixed at E = 1/σ then ϕE ( μ ) = exp{−μ/σ} which recovers the classical R 0 c [eq ( 1 ) ] whereas assuming an exponentially distributed EIP ( ϕE ( μ ) = σ/ ( σ + μ ) ) recovers the alternative R0 [eq ( 2 ) ] . A key point is that the generalised predictions for the reproductive number R 0 g depends on the MGFs of the EIP and gonotrophic cycle . MGFs depend upon the full distribution , therefore R 0 g can be estimated using all the information from an entomological study rather than just summary statistics such as mean ( or variance ) of gonotrophic cycle duration . In Results we will concentrate on gamma distributed durations which are determined by their mean and dispersion around the mean , however there is no theoretical necessity to use these distributions in general . Each vector has an age a , which defines the time since its first gonotrophic cycle was initiated after the end of its juvenile period . After reaching adulthood the vector has a random exponentially distributed lifetime L with mean E [ L ] = 1 / μ described by the probability density function , f L ( a ) = μ exp { - μ a } . ( 12 ) Exponentially distributed lifetimes , which derive from a constant hazard , have the advantage of a memoryless property . Having survived to some age a the vector’s expected remaining life is still 1/μ; that is the age of vector doesn’t effect its hazard of dying . Since we acknowledge that the age of the vector effects its biting rate we necessarily consider the age distribution of the vector population . In principle the vector age distribution could be seasonally varying , however we focus on the age distribution at population equilibrium . The age of a vector selected at random from its population at equilibrium is a random variable A . The equilibrium age probability density ( fA ( a ) ) can be calculated using the microcosm principle for population processes [20] which states that the proportion of a population in some state is proportional to the mean lifetime an individual would spend in that state . In this case , the microcosm principle gives , f A ( a ) = P ( L > a ) E [ L ] = μ exp { - μ a } . ( 13 ) For exponentially distributed life-times , the equilibrium population density of vector age is equal to the life duration density i . e . fL = fA , but this is not generally true . After successful inoculation with the pathogen the vector will become infectious after a random EIP duration E with density function fE . The memoryless property of the vector lifetime distribution implies that the probability of surviving its EIP ( PE ) is given by the MGF of the EIP duration evaluated at the vector mortality rate: P E = ℙ ( L > E ) = ∫ 0 ∞ ∫ 0 ∞ 1 ( x > y ) μ exp { − μ x } f E ( y ) dxdy = ∫ 0 ∞ exp { − μ y } f E ( y ) dy = ϕ E ( μ ) . ( 14 ) Where 1 ( . ) is an indicator function , taking the value 1 if the statement is true or 0 otherwise . Since exp{−μx} is a convex function in x then Jensen’s inequality gives that for any distribution of E ( c . f . inequality ( 2 . 23 ) in Klebaner [21] ) : P E ≥ exp { - μ E [ E ] } = exp { - μ / σ } . ( 15 ) Eq ( 15 ) shows that the probability of an inoculated vector surviving a fixed EIP is the lowest possible estimate for any EIP distribution . Vector biting follows a natural repeating cycle: the vector finds a host to bite at the end of each of its gonotrophic cycles and , once successfully fed , initiates a new gonotrophic cycle which includes locating a suitable oviposition site , oviposition , oogenesis and then subsequent host seeking . We model the gonotrophic cycle durations as a collection of independent and identically distributed ( i . i . d ) random variables labelled Gn , each with a common density function , fG . The ( random ) number of bites the vector has managed by age a , denoted B ( a ) , is the same as the number of gonotrophic cycles completed , B ( a ) = max n : ∑ k = 1 n G k ≤ a . ( 16 ) Processes such as this that count the number of i . i . d . waiting periods completed up to a given time are known as renewal processes , and their probabilistic properties have been extensively analysed [15] . For vectors dying at constant rate μ , the chance of surviving each gonotrophic cycle is independent of the number of cycles survived in the past . Eq ( 14 ) gives the probability of a vector surviving its EIP , by exactly the same argument the vector survives each gonotrophic cycle with probability ϕG ( μ ) up until its death therefore the lifetime number of bites by a vector is distributed geometric , with mean ϕG ( μ ) ( 1 − ϕG ( μ ) ) −1 . The Poisson process , which is associated with constant rates and hence ODE models , is an important special case of a renewal process . The implicit assumption that the average biting rate per vector is some fixed rate α , independent of the age and time since last bite of the vector , is equivalent to assuming that each gonotrophic cycle is exponentially distribution with average duration E [ G ] = 1 / α . However , we do not need to restrict our attention to this special case . Over a long life duration the elementary renewal theorem [15] gives that the expected bites per unit time will converge on the inverse of the average duration α , lim a → ∞ E [ B ( a ) ] a = α . ( 17 ) Eq ( 17 ) potentially explains why it is uncommon in the mathematical epidemiology literature to model epidemic contact processes as general renewal processes , as opposed to Poisson processes , since if the life time of an individual is long compared to its inter-contact periods a Poisson process is a reasonable approximation . However , for short lived vectors this is not true; the average biting rate per vector will crucially depend upon the age distribution amongst the population . We can now write the average biting rate , denoted α ( a ) , for vectors at age a in terms of probability densities: α ( a ) = ∑ n ≥ 1 f G * n ( a ) . ( 18 ) Where f G * n is the density function for the sum of gonotrophic cycles ∑ k = 1 n G k ( see supporting information S1 Text ) . The equilibrium population average biting rate per vector is therefore , M E [ α ( A ) ] = M μ ∫ 0 ∞ exp { − μ a } α ( a ) da = M μ ∑ n ≥ 1 [ ϕ G ( μ ) ] n = M μ ϕ G ( μ ) 1 − ϕ G ( μ ) . ( 19 ) Here we have used standard results on the MGF of sums of independent random variables and geometric sums . The equilibrium biting rate E [ α ( A ) ] differs from the long term biting rate α because of the non-exponentially distributed gonotrophic cycles and the typically short lives of the vector , such that many vectors may die before they seek their first blood meal ( Fig 2 ) . Note that Eq ( 19 ) implies that in the long life limit E [ α ( A ) ] → α irrespective of the gonotrophic cycle distribution , and hence results are independent of the assumed distribution . Vectors that become inoculated via a bloodmeal from an infectious host have bitten at least once , therefore the population equilibrium biting rate is not appropriate to model their subsequent biting process . The expected number of future bites by an inoculated vector after its EIP , denoted BI , can be calculated directly as , E [ # bites after EIP by inoculated vector ] = B I = ∫ 0 ∞ P ( L > t ) P ( E ≤ t ) α ( t ) dt . ( 20 ) That is the expected infectious bites from an inoculated vector are due to the aggregate of the biting rate at all times t after the inoculating bite , weighted by both the probability that the vector is still alive at that time and has completed its EIP ( Fig 3 ) . Eq ( 20 ) is numerically solvable using that α ( • ) is the rate of change of the solution to the renewal equation with waiting time distribution G ( see supporting information S1 Text ) . Probabilistically , the exact value for BI is difficult to compute because the probability of the vector surviving from the end of its EIP until its next bite depends upon the distribution of E in a complex fashion . However , it is clear that only the small proportion of vectors that are by chance comparatively long-lived contribute to BI when the EIP is typically longer than life expectancy . Therefore it can be argued that the asymptotic biting rate α ( eq 17 ) is appropriate for such vectors conditional on having survived their EIP , i . e . B I ≈ P E α μ = ϕ E ( μ ) α μ . ( 21 ) Eq ( 21 ) is exact when the gonotrophic cycle durations are exponentially distributed . In the supporting information we make the argument for Eq ( 21 ) as a general approximation to Eq ( 20 ) more rigorous by considering the long-time distribution of remaining time after the EIP until next bite . We also consider the special case of each gonotrophic cycle being of fixed , rather than random , duration ( S1 Text ) .
Using a gamma distribution to interpolate between exponentially distributed ( when the gamma shape parameter is 1 ) and constant periods ( when the gamma shape parameter diverges to infinity with fixed gamma mean ) is a common modelling choice due to its comparative simplicity and flexibility with a pedigree in epidemic modelling [28] . The dispersion of a gamma distributed random variable X , dX , is defined as , d X = VAR ( X ) E [ X ] 2 . ( 23 ) The dispersion of the random period X interpolates between fixed ( dX = 0 ) and exponential ( dX = 1 ) , it is also possible for a duration to be over-dispersed ( dX > 1 ) but we restrict attention to under-dispersal in this work . The dispersion is also usefully incorporated into the MGF of a gamma distributed random variable X: ϕ X ( μ ) = ( 1 + μ E [ X ] d X ) - 1 / d X . ( 24 ) Modelling E and G as gamma distributed with dispersions , dE and dG respectively allows us to calculate the population average biting rate per host and probability of surviving its EIP as M E [ α ( A ) ] = M μ 1 + μ d G α 1 / d G - 1 - 1 , ( 25 ) and P E = 1 + μ d E σ - 1 / d E ( 26 ) respectively . The mortality rate , mean gonotrophic cycle duration and mean EIP duration for BTV are all strongly temperature dependent for C . sonorensis with the reproductive ratio being maximal at around 20–25°C [9] . At 20°C , using realistic parameters , only a minority of midges are expected to survive the average EIP duration of BTV; the chance long-lived tail of the midge population are the sole contributors to the outbreak . However , the actual survival probability for the EIP varies from the 15 . 6% survival ( dE = 0; the classical assumption that all EIP durations are identical ) to 35 . 0% survival ( when dE = 1 ) . The population averaged biting rate per midge is similarly sensitive , and varies from 0 . 18 bites per midge per day ( dG = 1; the classical assumption that the gonotrophic cycle is exponentially distributed ) to 0 . 09 bites per midge per day ( dG = 0 ) ( Fig 2 ) . Turning our attention to the relative reproductive ratio ( Rrel ) , and using the numerically exact form of BI ( eq ( 20 ) and Fig 3 ) , we find that this increases with increasing dispersion of either distribution . As such , the reproductive ratio is maximised by both distributions being exponential ( Rrel = 2 . 24 ) and is lowest when both are fixed ( Rrel = 0 . 50 ) ( Fig 3 ) . The classical R 0 c for BTV can therefore be either a significant overestimate or a significant underestimate compared to the true prediction that accounts for dispersion . The approximate R ˜ 0 g is exact when the midges are expected to bite according to a Poisson process ( dG = 1 ) . Whenever the gonotrophic cycle duration has high dispersal ( dG > 0 . 5 ) the approximate reproductive ratio is a good estimate across the entire EIP dispersion range ( Fig 3 ) . However R ˜ 0 g can be both an underestimate or an overestimate of R 0 g when dG < 0 . 5 depending on an complex interplay between the distributions of E and G . In particular , when the EIP duration is nearly exponential ( dE ≈1 ) and the gonotrophic cycle is under-dispersed ( dG < 0 . 5 ) , with the worst relative error being R ˜ 0 g / R 0 g = 1 . 21 when ( dE = 1 , dG = 0 ) . This can be understood through the failure of the assumption that only chance long-lived midges with biting rates well approximated by the asymptotic rate α survive their EIP . The most important feature of the reproductive ratio as a measure of transmission intensity is that it yields a threshold quantity for disease persistence , and hence provides a quantitative insight into control . If an agency tasked with disease control can cause the reproductive ratio to fall below unity for an extended period of time then eradication can be achieved locally . One method for reducing the reproductive ratio is by vaccinating the host population; in the case of BTV there was a mass vaccination of northern European cattle during 2008 which was ultimately effective at eliminating BTV incidence in northern Europe . However , a vaccine for a novel or unexpected strain might not be immediately available , as occurred in northern Europe in 2006 which was invaded by BTV-8 serotype rather than the BTV-2 or BTV-4 serotypes which circulate in southern Europe [17] , in which case other control measures need to be considered . Reducing the midge population by using insecticides would seem to be an obvious solution , if this could be done whilst respecting standards of user and environmental safety . Wide-spread application of insecticide around a farm has not been recorded as successfully causing significant local midge population reduction [29] , but it has been suggested that targeting spraying of an insecticide with long residual life in proximity to cattle might be more successful [30] . To model the effect of such insecticide spraying proximate to cattle we assume that the midge mortality is increased by an excess mortality μe but the overall midge to host ratio ( M ) remains unchanged , due to the inability to target the entire midge population . If we make the classical assumptions ( dE = 0 , dG = 1 ) , then the critical excess mortality ( μ c * ) required to achieve eradication can be determined as the solution to: M V α 2 γ ( μ + μ c * ) exp ( - ( μ + μ c * ) / σ ) = 1 . ( 27 ) This is derived from eq ( 1 ) , by including the excess mortality and insisting that the resultant reproductive ratio is one . For disease control using vaccination there is a simple relationship between the critical vaccination coverage of hosts and the reproductive ratio; such that the critical vaccination coverage amongst hosts implied by a classical estimate R 0 c is 1 - 1 / R 0 c . For eradication via increasing midge mortality the relationship between the critical excess mortality and the reproductive ratio estimate R 0 c derived using the classical assumptions ( e . g . [26] ) solves , μ μ + μ c * exp { - μ c * / σ } = 1 R 0 c . ( 28 ) Note that μ c * depends on the reproductive ratio estimate R 0 c , the midge mortality rate μ and the incubation rate σ in contrast to the critical vaccination coverage of hosts which depends only on R 0 c . A natural consideration is the effect that generalising the distribution of the EIP and the gonotrophic cycle durations has upon the predicted critical excess mortality or critical vaccination coverage needed for eradication . We begin by considering excess mortality and assume that we have access to a classical estimate of the pre-control reproductive ratio R 0 c . In the generalised setting , the midge population biting rate depends on the precise nature of the excess mortality so we consider two separate scenarios: Where B I ( μ + μ e * ) is the expected number of bites after its EIP by a midge undergoing excess mortality μ e * . It should be noted that Eqs ( 29 ) and ( 30 ) are defined in terms of an estimate of the reproductive ratio based on classical assumptions , which can be related to a generalised reproductive ratio via Eq ( 22 ) . For either control scenario calculating an approximate critical excess mortality ( μ ˜ e * ) by using the approximate expression for BI Eq ( 21 ) is significantly easier numerically . The quality of the analytic approximation to the critical excess mortality has similar dependency on the dispersion parameters as the approximate reproductive ratio R ˜ 0 g: the approximation is broadly good when dE < 0 . 5 or when the gonotrophic cycle is more dispersed than the EIP ( S1 Fig ) . The classical formulation of the reproductive ratio Eq ( 1 ) suggests that increasing the mortality rate of vectors could be an effective control strategy since R 0 c decreases faster than exponentially with increasing μ . In the generalised setting this is not necessarily true; the sensitivity of R 0 g to μ can be sub-exponential , implying that increasing vector mortality is less effective at reducing the proportion of inoculated vectors expected to survive their EIP than the classical prediction would predict . On the other hand , in the generalised setting the mean vector population biting rate can decrease with increasing μ; an effect that reduces transmission and which is absent from the classical Ross-Macdonald model . In fact , the generalised critical value , μ e * , can be either greater or less than the value implied by the classical assumptions ( μ c * ) derived from solving Eq ( 28 ) , with the relative critical excess mortality ( μ r e l * = μ e * / μ c * ) increasing for both increasing dG and dE ( Fig 4 ) . Unsurprisingly , the parameter region where control using spraying is expected to be harder than the classical prediction ( i . e . μ r e l * > 1 ) is larger for scenario 2 , which is more pessimistic about the efficacy of insecticidal spraying than scenario 1 ( Fig 4 ) . Because the critical vaccination coverage can be expressed as depending only on the reproductive ratio , the curve Rrel = 1 separates the region where it is more difficult from the region where it is less difficult to eradicate using host vaccination than the classical prediction . However , the Rrel = 1 curve is not identical to the curve defined by μ r e l * = 1 ( Fig 4 ) . This has an important consequence , even if R 0 g = R 0 c ( i . e . the critical host vaccination coverage is the same even when accounting for the distributions ) then the critical amount of excess mortality that insecticide based control must achieve in order to eradicate may be different depending on the distributions . In fact , for culicoides biting midges the amount of necessary insecticide control predicted using classical assumptions is an underestimate everywhere along the Rrel = 1 level curve for scenario 1 , and the underestimation is even more dramatic for scenario 2 . In fact , whenever the generalised reproductive ratio and the classical reproductive ratio coincide , the classical model is always overly optimistic about the how little excess mortality will be required to achieve eradication . This is true for both varying ( dG , dE ) for a BTV epidemic with the classical estimate R 0 c = 2 and when fixing dE = 0 . 130 ( the median posterior estimated dispersion for the EIP of C . bolitinos infected with BTV-1 serotype [19] ) allowing ( R 0 c , d G ) to vary ( Fig 4 ) . However , it should be noted that there is an appreciable region of parameter space where eradication is easier to achieve than expected using classical modelling assumptions . If dG is sufficiently small then in most cases we predict that it could be easier to eradicate via increasing adult vector mortality than predicted by classical modelling . Whether the findings of this paper are optimistic or pessimistic with regard to insecticidal spray efficacy will ultimately be resolved by more detailed data on biting vector gonotrophic cycle dispersion .
In the model presented here we assume that vector mortality is due to a constant background hazard rate μ and the vectors’ first gonotrophic cycle after emergence as an adult is similar to subsequent cycles . An alternative is to explicitly include heightened mortality risk associated with vector activities such as host-seeking or oviposition as has been done in modelling papers of mosquito-borne diseases ( e . g . [31] ) . Introducing additional risk of vector mortality immediately before and after biting modifies the time-dependent biting rate α ( • ) by incorporating the chance that biting leads to vector death . α ( • ) is the solution to a modified renewal equation and therefore introducing this extra model component represents no extra difficulty compared to only having background mortality ( see S1 Text ) . Many midge species are autogenous: their females emerge as adults ready to produce and oviposit an initial egg batch without a bloodmeal [32] . From an midge ecology perspective this allows the midge population to persist even when hosts are frequently absent . Therefore , from an epidemiological perspective the waiting time until the first bite from a female autogenous midge on a host occurs at the end of the second gonotrophic cycle . This possibility can be included in our model since we base our analysis on renewal theory: renewal processes with an initial waiting period that is different from subsequent waiting periods ( known as delayed renewal processes [15] ) are theoretically standard ( S1 Text ) . We did not explore these biologically motivated modelling extensions further due to our focus on bluetongue diseases spread by Culicoides genus biting midges; current best estimates of midge mortality in the wild are based on estimates of the proportion of parous midges in the population [13] . This gives an estimate of the mortality rate based on the probability of surviving a gonotrophic cycle which cannot be disentangled into separate hazards . Moreover , no autogenous midge species has been implicated in the transmission of any arbovirus , including bluetongue [32] . The classical Ross-Macdonald theory of VBDs created a critically important scientific framework with a measure of generational transmission intensity ( R0 ) at its heart . However the specific practical insights derived from the Macdonald reproductive ratio ( for example the implied efficacy of increasing the mortality rate of adult insect vectors in reducing transmission ) are undermined by the growing realisation amongst epidemiologists of mosquito-borne diseases that different methods for estimating R0 produce strongly differing results ( cf . Smith et al for a discussion on the problems associated with validation of the mathematical theory of mosquito-borne pathogens [6] ) . The problems with theoretic cross-validation appear to have not been explicitly stated in the literature of midge-borne pathogens but the issues are essentially similar . In summary , it has become generally accepted in the modelling community that , whilst all models are simplifications of reality , the classical simplifications of Ross-Macdonald theory may produce misleading results ( cf . recent reviews [5 , 6] ) . The development of mathematical models of VBDs that deviate from Ross-Macdonald theory has generally focused on heterogeneity in transmission , whether due to heterogenous distribution of vectors across space , for example the recent northern European outbreak of BTV has inspired a number of explicitly spatial models ( for example [33–35] ) , or because of the observation that ( typically ) a disproportionately greater number of bites will be distributed across a small number of vertebrates ( as discussed in [26] ) . It seems that questions about the role of chance in generating variation in life histories between vectors , even if they are otherwise identical from a modelling perspective , has been largely overlooked despite being in principle empirically observable in the laboratory setting . In this work , we have demonstrated that relaxing the classical assumptions on the random distributions underlying the calculation of a reproductive ratio for a VBD can have a significant impact on the prediction , even without additional modelling of heterogeneous biting or spatial vector distribution . The impact of the precise distribution of life history events is accentuated for short-lived vectors; thus the economically important vector-disease complex of BTV spread by , the typically short-lived , culicoides genus biting midges is an ideal case study for our theory . Our main results are presented in terms of moment generating functions of fairly arbitrary period distributions , so that any reasonable distribution can be used for modelling aspects of the vector life histories . However , our illustrative examples are calculated for the gamma distribution , a natural choice since it interpolates between the constant duration and the exponentially distributed duration . Gamma distributions have been used extensively to model the durations of latency ( EIP ) and viraemia , most commonly as a discrete sum of exponential distributions ( a special case of the gamma distribution called the Erlang distribution ) , in both the vector-borne disease literature [9] and the directly transmitted disease literature [28] . However , the timing of epidemiological relevant contacts , ( e . g . successful bites ) are overwhelmingly modelled as a Poisson process throughout the mathematical modelling literature . We show that the generalisation of the contact model from a Poisson process to a renewal process , in conjunction with the typically short lives of the vector and a general EIP duration , has significant effects on the basic predictions of an otherwise standard VBD model . The more general modelling of the vector’s biting process as a renewal process , and the EIP duration as non-constant , predicts that the classical reproductive ratio can be either a significant overestimate or underestimate compared to the generalised prediction . Moreover , the generalised modelling approach cannot be disentangled from considerations of the age distribution of the vector population . However , by assuming that the population is at equilibrium and making theoretically well-motivated approximations ( as well as assuming constant mortality rates ) we are still able to present the dependence of R0 on the entomological situation in an explicit manner . For the motivating case of midges spreading BTV , the reproductive ratio is increasing with increasing variation in either the latency ( EIP ) duration or the time between bites ( gonotrophic cycle ) . This can be understood by considering two consequences of the short life-span of the ‘typical’ midge: firstly , a significant proportion of midges can die whilst their expected biting rate is low ( due to relaxing the Poisson biting process assumption ) and secondly , that the ‘typical’ midge inoculated with BTV will not survive her EIP ( which can depend sensitively on the EIP distribution ) . Only the ‘tail’ of the inoculated midge population actively contribute to BTV incidence amongst hosts , and therefore random variation between vector life history events becomes a crucial factor . In particular , the potential importance of the precise details about the distribution of the gonotrophic cycle duration seems to be under-recognised in both the theoretical and entomological literature . These findings echo results from at least one other model which generalises the vector biting process [36] and suggests that greater effort must be devoted to understanding and quantifying the detailed vector biology both at the individual and population levels in order to accurately parametrise these kinds of models . The popularity of R0 as a metric of transmission intensity is in a large part due to its uncomplicated relationship with disease eradication: if R0 can be reduced below unity for a sufficient period of time then the disease must be eradicated . The generalised prediction of R0 presented in this work therefore leads to a reassessment of the efficacy of standard control measures such as the mass vaccination of hosts or the insecticidal spraying against adult vectors . The predicted critical vaccination coverage amongst hosts can still be expressed as only depending on the estimate of the reproductive ratio in the standard fashion , however the critical excess mortality amongst vectors that must be achieved depends on both the estimate of the reproductive ratio and other entomological/epidemiological factors such as basic vector mortality rate and the incubation rate for the pathogen even when making the classical Ross-Macdonald assumptions about the VBD . Predicting the critical excess mortality in the more general setting is further complicated by the necessity to consider the effect that excess mortality has on the age distribution of the vector population . Trials of the ability of insecticidal spraying aimed at adult midges to reduce the local midge population have been pessimistic , leading to speculation that much more must be known about the resting places of adult midges in order to design better spraying protocols [29] . We would add to this discussion further speculation upon which part of the midge population is affected by the additional environmental hazard of a persistent insecticide . Does insecticidal spraying reduce biting from the susceptible mass of the vector population due to shifting their age profile ( our scenario 1 ) , or not ( our scenario 2 ) ? In either scenario we find that even when the reproductive value predicted by the generalised model agrees with its classical counterpart then the classical model still under-estimates how much excess mortality must be achieved in order to eradicate . None of the arguments above should be read as implying that insecticidal spraying should not be used as a control measure against a vector-borne disease in any circumstance . It might well be the case that a vaccine for hosts might be unavailable or prohibitively expensive whereas a suitable , and safe , insecticide might be readily available and cheap . An intriguing possibility is to target the older vectors that make up the tail of the age profile using late life acting insecticides ( LLAIs ) . This has been suggested for controlling malaria so that evolutionary pressure on the mosquito population to develop insecticidal resistance is sharply diminished [37] . Modelling the EIP as random rather than fixed always favours vector survival of EIP [eq 15] . This implies that more ‘young’ inoculated vectors will be actively infectious than expected by a model with a fixed EIP; a pessimistic finding with respect to the efficacy of LLAIs whether the LLAI is effective after a delay since exposure or simply preferentially affects older vectors . An advantage of the modelling approach taken in this work is that the action of the LLAI can be explicitly age-dependent or dependent upon real time since exposure . In Read et al [37] the action of the LLAI on the vector is modelled as occurring after a number of vector gonotrophic cycles . However , the age of a vector after a number of gonotrophic cycles is uncertain due to variation in the cycle duration . Therefore , experimental data on age-dependent ( or time delayed ) LLAI efficacy might be difficult to interpret in the context of a model which focuses only on the number gonotrophic cycles a vector survives . Our modelling approach doesn’t suffer from this problem and experimental data on LLAIs could be integrated into our model directly . This would certainly be an interesting direction for future research . Mathematical models are now commonly used to underpin policy decisions concerning disease control in both developed and low- and middle-income countries . We have used BTV which affects livestock spread by Culicoides genus midges as a motivating example . However , the methodology could equally apply to Leishmania in humans spread by sandflies , which have similar entomological characteristics to midges . Similarly , while mosquitoes do not necessarily match all assumptions within our model , mosquito-borne diseases could be accommodated with only a slight increase in model complexity . Our work has highlighted the sensitivity of model predictions to entomological and epidemiological details . This points to further avenues of applied experimental or observational research: 1 ) the acquisition of more detailed knowledge concerning the life cycles of potential vectors , in particular going beyond the measurement of simple averages to include variability; 2 ) studies of the life expectancy and age profiles of vector populations in the wild; 3 ) a more comprehensive investigations of the response of vector populations to insecticide-based control measures . Developing highly informative policy-relevant prediction for the future control of vector-borne diseases is likely to therefore require a combination of state-of-the-art models , with meticulous quantitative studies of the insect vector .
|
The basic reproductive ratio ( R0 ) is a crucial measure of transmission intensity , lying at the interface between mathematical modelling and policy decision making . If control measures can induce a situation where R0 ≤ 1 for a sustained period of time then the pathogen must be eradicated . For diseases spread by short-lived insect vectors a modeller can not calculate R0 without addressing questions of chance such as , “What percentage of vectors will survive their extrinsic incubation period ( EIP ) to become infectious ? ” . Classical Ross-Macdonald theory provides answers for the modeller by making certain concrete assumptions , such as a fixed length EIP and exponentially distributed times between vector blood-meals . Using bluetongue virus spread by biting midges as an exemplar we demonstrate that biologically plausible alterations to the classical assumptions can significantly change the modeller’s prediction of R0 with both serious over-estimation and under-estimation being possible . The important modelling/control question , “How does R0 respond to increased vector per-capita mortality ? ” , is also found to depend strongly on details of the vector life-cycle expressed in the language of probability distributions .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"death",
"rates",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"infectious",
"disease",
"epidemiology",
"demography",
"pathogens",
"immunology",
"vector-borne",
"diseases",
"culicoides",
"microbiology",
"animals",
"reoviruses",
"bluetongue",
"virus",
"viruses",
"preventive",
"medicine",
"rna",
"viruses",
"population",
"biology",
"infectious",
"disease",
"control",
"vaccination",
"and",
"immunization",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"agrochemicals",
"medical",
"microbiology",
"epidemiology",
"microbial",
"pathogens",
"insects",
"agriculture",
"arthropoda",
"people",
"and",
"places",
"insecticides",
"population",
"metrics",
"viral",
"pathogens",
"biology",
"and",
"life",
"sciences",
"organisms"
] |
2016
|
The Interaction between Vector Life History and Short Vector Life in Vector-Borne Disease Transmission and Control
|
Biological rhythms play a fundamental role in the physiology and behavior of most living organisms . Rhythmic circadian expression of clock-controlled genes is orchestrated by a molecular clock that relies on interconnected negative feedback loops of transcription regulators . Here we show that the circadian clock exerts its function also through the regulation of mRNA translation . Namely , the circadian clock influences the temporal translation of a subset of mRNAs involved in ribosome biogenesis by controlling the transcription of translation initiation factors as well as the clock-dependent rhythmic activation of signaling pathways involved in their regulation . Moreover , the circadian oscillator directly regulates the transcription of ribosomal protein mRNAs and ribosomal RNAs . Thus the circadian clock exerts a major role in coordinating transcription and translation steps underlying ribosome biogenesis .
Circadian rhythms in behavior and physiology reflect the adaptation of organisms exposed to daily light-dark cycles . As a consequence , most aspects of metabolism and behaviour are under the control of these rhythms [1] . At a molecular level , in all the studied species , the rhythmic expression of the genes involved originates in the network of interconnected transcriptional and translational feedback loops [2] . In mammals , the heterodimer composed of BMAL1 and its partners CLOCK or NPAS2 is a transcriptional activator that regulates transcription of the Period ( Per ) and Cryptochrome ( Cry ) genes that code for repressors of BMAL1 heterodimer activity , thus closing a negative feedback loop that generates rhythms of approximately 24 h [1] , [2] . Many efforts during the last decade have characterized rhythmically expressed genes and delimit the impact of the circadian clock on physiology . Numerous circadian transcriptome studies in different species and organs show that approximately 10% of the genes are rhythmically expressed . The functions of these genes established the role of the circadian clock in temporally gating rhythmic physiology [1] , [3] . However , increasing evidence suggests that transcriptional mechanisms are not sufficient to explain numerous observations . For example , it has been shown that many oscillating proteins in mouse liver are encoded by constantly expressed mRNAs [4] . Interestingly , among the rhythmically expressed genes in the liver , we noticed the presence of several genes encoding proteins involved in mRNA translation , including the components of the translation pre-initiation complex 5 , 6 . In its inactive state , this complex is composed of the mRNA cap-binding protein eukaryotic translation initiation factor 4E ( EIF4E ) bound to the hypophosphorylated form of EIF4E-binding protein ( 4E-BP ) that acts as a translational repressor . Upon stimulation , phosphorylation of 4E-BP releases EIF4E , which can then interact with the scaffold protein eIF4G and the rest of the EIF4F complex ( EIF4A , EIF4B , and EIF4H ) to initiate translation [7] . We therefore investigated whether the circadian clock might coordinate translation in mouse liver . Here we indeed show that the circadian clock controls the transcription of translation initiation factors as well as the rhythmic activation of signaling pathways involved in their regulation . As a consequence , the circadian clock influences the temporal translation of a subset of mRNAs mainly involved in ribosome biogenesis . In addition , the circadian oscillator regulates the transcription of ribosomal protein mRNAs and ribosomal RNAs . These results demonstrate for the first time the major role of the circadian clock in ribosome biogenesis .
We investigated whether the circadian clock might coordinate translation in mouse liver . Indeed , quantitative reverse transcription ( RT ) -PCR analyses confirmed that mRNAs of most of the factors involved in translation initiation are rhythmically expressed with a period of 24 h ( Figure 1A; statistical analyses are given in Table S1 ) . Interestingly , while we did not observe any significant variations in protein abundance , rhythmic phosphorylations were strongly manifested during two consecutive days , emphasizing the robustness of these rhythms ( Figure 1B; quantification and statistical analyses of the data are given on Figure S1 and Table S2 ) . EIF4E is mostly phosphorylated during the day , with a peak at the end of the light period ( ZT6-12 ) , whereas EIF4G , EIF4B , 4E-BP1 , and ribosomal protein ( RP ) S6 ( RPS6 ) are mainly phosphorylated during the night , which is , in the case of nocturnal animals like rodents , the period when the animals are active and consume food . Phosphorylation of these factors is well characterized and involves different signaling pathways [8] whose reported activity perfectly correlates with the observed phosphorylation rhythm . EIF4E is phosphorylated by the extracellular signal-regulated protein kinase ( ERK ) /mitogen-activated protein kinase ( MAPK ) -interacting kinase ( MNK ) pathway [9] , which is most active during the day , at the time when EIF4E reaches its maximum phosphorylation ( Figure 2A; quantification and statistical analyses of the data are given on Figure S2 and Table S2 ) . On the other hand , EIF4G , EIF4B , 4E-BP1 , and RPS6 are mainly phosphorylated by the target of rapamycin ( TOR ) complex 1 ( TORC1 ) [10] , which is activated during the night , at the time when the phosphorylation of these proteins reaches its maximum level . TORC1 , in turn , is negatively regulated by the tuberous sclerosis protein complex ( TSC ) , whose activity is under the control of the phosphoinositide 3-kinase ( PI3K ) /AKT , ERK , and the energy sensing 5′ adenosine monophosphate-activated protein kinase ( AMPK ) pathways [10] , [11] . As reported [12] , AMPK is active during the day and mediates the activation of TSC2 , contributing to the repression of TORC1 in the period of energy and nutrient restriction . Conversely , during the night , TORC1 is activated probably through TSC2 inhibition by PI3K via TORC2 [13] . Interestingly , we found that mTor , its partner Raptor , as well as its regulating kinase Map3k4 , are also rhythmically expressed , thus potentially further contributing to the rhythmic activation of TORC1 ( Figure S3; Table S1 ) . ERK is activated during the day in synchrony with the rhythmic expression of Mnk2 ( Figure S3 ) , contributing to EIF4E phosphorylation during this period . However , its downstream target RPS6 Kinase ( RSK ) seems to contribute only marginally to the phosphorylation of RPS6 in mouse liver ( Figures 1B and 2A ) . The rhythmic phosphorylation of 4E-BP1 resulted in its release from the mRNA cap-mimicking molecule 7-methyl-GTP from ZT14 to ZT22 ( Figure 2B; Table S2 ) , allowing the rhythmic assembly of the EIF4F and potentially mRNA translation . The rhythmic expression of mRNA encoding translation initiation factors , TORC1 complex component , and a kinase activating these factors is independent of light as it is maintained under constant darkness , even if the phase seems to be advanced ( Figure S4A ) . Interestingly , activation of the TORC1 pathway is also maintained under constant darkness but with an advanced phase ( Figure S5A ) . Since nutrient availability is a potent activator of the TORC1 pathway [13] , we asked whether these parameters are also rhythmic under conditions of starvation . We found that expression of mRNA encoding translation initiation factors , TORC1 complex component , and a kinase activating these factors is still rhythmic under starvation ( Figure S4B ) , even when this starvation occurs under constant darkness ( Figure S4C ) . This result unambiguously demonstrates the role of the circadian clock in the expression of these genes . In addition , phosphorylations of RPS6 and 4E-BP1 are still rhythmic under starvation , whether or not the mice are under a light-dark regimen or in constant darkness ( Figure S5B and S5C ) , confirming previously published observations [14] . Interestingly , TORC1 activation is in opposite phase with the clock-dependent rhythmic activation of autophagy in mouse liver [15] , a process inhibited by TORC1 but able to generate amino acids that can in turn activate TORC1 [16] . This might suggest that the circadian clock can regulate the two processes in a coordinated fashion . Importantly , rhythmic activation of TORC1 is not restricted to the liver as the same phosphorylation rhythm is found in kidney and heart , albeit with reduced amplitude ( Figure S6 ) . Meanwhile , TORC1 activation is constant in brain , lung , and small intestine , suggesting that the rhythmic nutrient availability due to the circadian clock-regulated feeding behavior is not sufficient by itself to explain the rhythmic activation of TORC1 . Diurnal binding of 4E-BP to EIF4E suggested that translation might be rhythmic in the liver . To test this hypothesis and to identify potential rhythmically translated genes , we purified polysomal RNAs , a RNA sub-fraction composed mainly of actively translated mRNA , every 2 h during a period of 48 h . We found that relative amount of this polysomal fraction follows a diurnal cycle , showing that a rhythmic translation does occur in mouse liver ( Figure S7 ) . This result confirms original observations based on electron microscopy and biochemical studies [17] , [18] . We therefore decided to characterize these rhythmically translated mRNAs through comparative microarray analysis of polysomal and total RNAs . While the obtained profiles in polysomal and total RNAs fractions are highly similar for most mRNAs ( examples of rhythmic mRNAs are given on Figure S8 ) , 249 probes showed a non-uniform ratio in diurnal polysomal over total mRNAs ( Figure 3A ) . This means that approximately 2% of the expressed genes are translated with a rhythm that is not explained by rhythmic mRNA abundance as in most cases , the total mRNA levels were constant while the polysomes-bound mRNA levels fluctuated during the 24-h cycle ( Figures 3B and S9 ) . Among translationally regulated genes , 70% were found in the polysomal fraction during the same time interval , starting at ZT8 before the onset of the feeding period and finishing at the end of the dark period ( Tables S3 and S4 ) . Most of these genes belonged to the 5′-terminal oligopyrimidine tract ( 5′-TOP ) family , known to be regulated by TORC1 [19] , but also by the level and phosphorylation state of EIF4E [20] , [21] . 5′-TOP genes are themselves involved in translation via ribosome biogenesis and translation elongation ( Table S4 ) . After confirmations of these results by quantitative RT-PCR ( Figure S10 ) , we wished to validate the periodicity in the amount of mRNAs purified in the different fractions obtain during polysomes purification over a 24-h period . Whereas a constitutively translated mRNA such as Gapdh is found all the time in the polysomal fraction ( with a small decrease in the middle of the light period when overall translation decreases ) , mRNAs coding for RPs are associated with the polysomal fraction only starting towards the end of the light period ( ZT8 ) and during the dark period ( Figure 3C ) . This result demonstrates a dynamic translation initiation of 5′-TOP mRNA starting before the onset of the feeding period , with a maximum at the beginning of the dark period . Next , we wanted to confirm that this rhythmic translation had an impact on the protein levels . With respect to RPs , while the half-life of mature ribosomes is approximately 5 d in rodent liver [22] , newly synthesized RPs have a half-life of only a few hours , as most of them are rapidly degraded after translation during the ribosome assembly process in the nucleolus [23] . We thus expected a rhythmic expression of this subpopulation of newly synthesized RPs in the soluble cytosolic fraction depleted of ribosomes after sedimentation . Indeed , under these conditions , RPs show a rhythmic abundance with highest expression during the night ( Figure 3D; quantification and statistical analyses of the data are given on Figure S11 and Table S2 ) . In some cases , we noticed a shallow decrease at ZT16-18 , potentially reflecting transport of RPs into the nucleolus for ribosome assembly . In addition to translational regulation , we also observed a diurnal expression of RP mRNAs , albeit with a small average peak to trough amplitude of approximately 1 . 2 . Taking into account their relatively long half-life ( 11 h ) [24] , we hypothesized that this minor fluctuation might reflect more pronounced rhythmic amplitudes in transcription as amplitude decreases with half-life [25] . In addition , it has recently been shown that the transcription of several RP mRNAs is directly controlled by the molecular oscillator in Drosophila head [26] . Indeed , pre-mRNA accumulation of several RP exhibited a rhythmic transcription , with an average amplitude of 3 . 5-fold with a maximum at ZT8 , just before the activation of their translation ( Figure 4A; statistical analyses are given in Table S1 ) . In addition , we found that the synthesis of the ribosome constituent precursor 45S rRNA is also rhythmic and synchronized with RP mRNAs transcription , indicating that all elements involved in ribosome biogenesis are transcribed in concert , then translated or matured . In yeast [27] and Drosophila [28] , transcription of RP mRNAs appears to be coordinated with rRNA transcription , which is a rate limiting step in ribosome biogenesis . On the other hand , in mammals , rRNA transcription is highly regulated by the upstream binding factor ( UBF ) , which establishes and maintains an active chromatin state [29] . Remarkably , we found that UBF1 is rhythmically expressed in mouse liver at both mRNA and protein levels ( Figure 4B; quantification and statistical analyses of the data are given in Figure S12A and Tables S1 and S2 ) , in phase with RP mRNAs and rRNAs transcription . In addition , rhythmic transcription of Ubf1 and Rpl23 genes is also independent of light and food ( Figure S4 ) . To test whether Ubf1 transcription is regulated by the circadian clock , we characterized its expression in arrhythmic Cry1/Cry2 knockout ( KO ) [30] and Bmal1 KO [31] mice , which are devoid of a functional circadian clock . Indeed , these mice do exhibit an arrhythmic pattern of activity under constant darkness , which is in general correlated with an arrhythmic feeding behaviour . As TORC1 , as well as other signaling pathways , are in part regulated by feeding through nutrient availability , we expect a temporally discontinuous and erratic activation of these pathways in the KO mice under unrestricted feeding . To verify this hypothesis , we measured activation of the TORC1 , AKT , and ERK pathways in Cry1/Cry2 and Bmal1 KO kept in constant darkness . As shown in Figure S13A , the rhythmic activation of these signaling pathways is indeed lost under this condition , confirming their arrhythmic activation . To highlight the role of the feeding regimen on this activation , we kept Cry1/Cry2 KO mice in constant darkness and sacrificed them at CT12 . We found a strong inter-individual variability in the activation of the TORC1 , AKT , and ERK pathways , reflecting the arrhythmic feeding rhythm of these animals ( Figure S13B ) . To circumvent this caveat and study the rhythmic translation in mice devoid of a functional molecular oscillator , we decided to place Cry1/Cry2 and Bmal1 KO under a light-dark regimen to keep a normal diurnal feeding behaviour due to masking . In addition , mice had access to food only during the dark phase to eliminate the effect of a potential disturbed feeding behaviour . Under these conditions , KO mice had a rhythmic feeding behaviour and thus potential differences in protein levels or pathway activity cannot be attributed to the arrhythmic feeding behaviour of these animals . We indeed found that UBF1 rhythmic expression is dependent on a functional circadian clock as it is impaired in both animal models ( Figure 4C and 4D; quantification and statistical analyses of the data are given in Figure S12B and Tables S5 , S6 , S7 , S8 ) . However , if UBF1 expression is persistently low in Cry1/Cry2 KO mice , this expression is constantly high in Bmal1 KO mice , suggesting the control of Ubf1 by a circadian clock-regulated transcription repressor . In addition , we observed that these animals lose also the synchrony and coordination of 45S rRNA and RP pre-mRNAs transcription ( Figures 5 , S14 , and S15; statistical analyses of the data are given in Table S5 and S6 ) . Indeed , decreased UBF1 expression in Cry1/Cry2 KO mice is correlated with lower 45S rRNA transcription , but higher and delayed RP pre-mRNAs transcription . Interestingly , Bmal1 KO mice present a complete arrhythmic transcription of RP pre-mRNAs , highlighting the crucial role of the circadian clock in the coordination of rRNA and RP mRNAs transcription . Rhythmic expression of genes coding for components of the translation initiation complex is strongly dampened or phase-shifted in both KO models , in addition to an altered level of expression ( Figures 5 , S14 , and S15; statistical analyses of the data are given in Tables S5 and S6 ) . However , we did not observe in general any significant variations in protein abundance , excepting a slight increase in EIF4E expression in Cry1/Cry2 KO mice , reflecting increased mRNA expression ( Figure 6A and 6C; quantification and statistical analyses of the data are given in Figures S16 , S17; Tables S7 and S8 ) . The variations in EIF4G levels reflect more the changes in its phosphorylation state , which regulates its stability [32] . While most of the signaling pathways are still rhythmic in Cry1/Cry2 KO mice , except for the ERK pathway and the downstream phosphorylation of EIF4E , which loses its rhythmic activation , the phase of the activation of the TORC1 and AKT pathways are advanced in comparison to wild-type ( WT ) mice ( Figures 6A and S16; quantification and statistical analyses of the data are given in Table S7 ) . As a consequence , the rhythmic expression of RPs is altered in Cry1/Cry2 KO mice ( Figure 6B; quantification and statistical analyses of the data are given in Table S7 ) , with an increased level of expression , likely because of the increased RP pre-mRNAs and EIF4E levels [20] , and a delayed phase of expression . Most of the rhythmic activation of the three pathways is also strongly altered in Bmal1 KO mice ( Figures 6C and S17; quantification and statistical analyses of the data are given in Table S8 ) . As shown in Figure 6D , the phase of RPs rhythmic expression is severely advanced with a maximum of expression in the middle of the day instead of the night ( Figure 6D; quantification and statistical analyses of the data are given in Table S8 ) .
The results presented here show that the molecular circadian clock controls ribosome biogenesis through the coordination of transcriptional , translational , and post-translational regulations . Moreover , the data strongly suggest that a functional molecular oscillator is required for a timely coordinated transcription of translation initiation factors , RP mRNAs , and rRNAs . The clock modulates the rhythmic activation of signaling pathways controlling translation through the TORC1 pathway , translation of RPs , and ribosome biogenesis ( Figure 7 ) . Interestingly , it has been reported that the size of the nucleolus , the site of rRNA transcription and ribosome assembly , follows a diurnal pattern with a maximum in the middle of the dark period [33] , which thus occurs in synchrony with the observed accumulation of RPs in the liver . The observed rhythmic ribosome biogenesis is substantiated by the previous observation showing that both size and organization of the nucleolus are directly related to ribosome production [34] . Remarkably , a coordinated rhythmic regulation of transcriptional and translational events for the biogenesis of ribosomes has also been suggested for the filamentous fungus Neurospora crassa [35] and for plants [36] , [37] . Since ribosome biogenesis is one of the major energy consuming process in cells [38] , its tight control is primordial to reduce interferences with other biological processes . In the case of mouse liver , we estimate that the decrease of translation during the light period is equivalent to 20% of the total translation ( Figure S7 ) , in agreement with previously published results [17] . Although moderate , this decrease affects translation of housekeeping genes like Gapdh ( Figure 3C ) and probably the translation of other genes . It means that the increase in ribosome biogenesis during the night could potentially influence the translation of many other mRNAs , however with a magnitude sufficiently low to not allow its detection by our method . Nevertheless , it is clear that this energy-consuming process has to be confined to a time when energy and nutrients are available in sufficient amount , which , in the case of rodents , is during the night period when the animals are active and consume food . Hence , all the elements required for translation have to be ready to start ribosome biogenesis during that time . This is achieved by increasing levels of rRNAs and RP pre-mRNAs just before the onset of the night , synchronized with the phosphorylation of EIF4E that increases 5′-TOP mRNAs translation [21] . Activation of the TORC1 pathway during this period promotes RPs synthesis , rRNAs maturation , and ribosome assembly . In addition activation of the ERK pathway correlates also with ribosome biogenesis [39] , strengthening the rhythmic nature of this process . Accordingly , orchestration of ribosome biogenesis by the circadian clock represents a nice example of anticipation of an obligatory gated process through a complex organization of transcriptional , translational , and post-translational events . As described in the introduction , the mammalian molecular circadian oscillator consists in interlocked feedback loops of transcription factors that generate a complex network of rhythmically expressed genes [3] . Within the core molecular clock , increasing evidence shows that post-translational modifications play a crucial role in the generation of circadian rhythms [40] . However , the circadian clock is also able to coordinate rhythmic post-translational activation of signaling pathways not directly involved in the molecular oscillator but rather in the sensing of the environment . The first described example consisted in the rhythmic activation of ERK in the suprachiasmatic nucleus ( SCN ) of the hypothalamus where the master circadian pacemaker is localized: if light stimulates ERK phosphorylation in the SCN in a time-dependent fashion , circadian ERK phosphorylation continues also in constant darkness , suggesting a crucial role of the circadian clock in this process [41] . Interestingly , the same observations have been made for the TORC1 pathway in the SCN [42] , [43] , and for the PI3K/AKT pathway in the retina [44] . Considering the fact that these two pathways have been recently identified as a potent regulators of circadian activity in Drosophila [45] , we expect that the role of the circadian clock-coordinated signaling pathways on circadian physiology will probably be emphasized in other organisms in the near future . With respect to rhythmic activation of signaling pathways in the liver , there are only few examples of such regulations . One example is the rhythmic activation of the PI3K/AKT pathway that is associated with food metabolism and rhythmic feeding behavior [46] . Recently , we also described a circadian clock-dependent rhythmic activation of the unfolded protein response regulating liver lipid metabolism [47] . In addition , it has been shown that the circadian clock is also able to regulate autophagy in mouse liver [15] . In this context , our discovery of the rhythmic ribosome biogenesis through coordination of the rhythmic activation of signaling pathways constitutes an important new element in this area of research . It has long been known that caloric restriction or intermittent fasting increases lifespan in a wide variety of models [48] . Increased lifespan has also been linked to the reduced activation of the TORC1 pathway , which , in turn , provokes a reduced mRNA translation [49] , [50] . The role of the TORC1 pathway in this translation-dependent extension of lifespan has been genetically confirmed in Caenorhabditis elegans [51] and Drosophila [52] , [53] . A similar scenario is also considered in mice since treatment with the TOR inhibitor rapamycin [54] or deletion of the TORC1 downstream protein kinase S6K1 [55] lead to increased lifespan . In addition , downregulation of various components of the EIF4F complex extends lifespan in C . elegans [56]–[59] , whereas inhibition of RPs genes expression extends lifespan in both Saccharomyces cerevisiae [60] and C . elegans [56] . Hence , keeping ribosome biogenesis , and translation in general , to their minimum levels plays a major role in the regulation of longevity [61] . Interestingly , all the genetically modified animal models presenting a disrupted circadian clock [62]–[64] or mice subjected to chronic jet lag [65] are subjected to premature aging and reduced lifespan . The deregulation of many other circadian-clock regulated processes can reduce life expectancy , like reduced xenobiotic detoxification [66] . We thus believe that the potential role of disorganized ribosome biogenesis on life expectancy , observed in animals devoid of a circadian clock , will be an exciting subject for further studies .
All animal studies were conducted in accordance with our regional committee for ethics in animal experimentation and the regulations of the veterinary office of the Canton of Vaud . C57Bl/6J mice were purchased from Janvier ( Le Genest ) or Charles River Laboratory ( L'Arbresle ) . Bmal1 floxed mice have been previously described [67] . These mice were crossed with mice expressing the CRE recombinase under the control of the CMV promoter [68] to obtain Bmal1 KO mice . Cry1/Cry2 double KO mice [30] in the C57Bl/6J genetic background have been previously described [69] . In all experiments , male mice between 10 and 12 wk of age are used . Unless noted otherwise , mice were maintained under standard animal housing conditions , with free access to food and water and in 12-h light/12-h dark cycles . However , for all experiments , animals were fed only at night during 4 d before the experiment to reduce effects of feeding rhythm . For experiments in constant darkness , mice were shifted into complete darkness after the last dark period and then sacrificed every 2 or 4 h during the next 48 h . For starvation experiments , mice were deprived from food during one complete night and then during the following 24 h , mice were sacrificed every 2 or 4 h . Livers were homogenized in lysis buffer containing 20 mM HEPES ( pH 7 . 6 ) , 250 mM NaCl , 10 mM MgCl2 , 10 mM DTT , 20 µg/ml cycloheximid , 10 U/µl RNase inhibitor , and a protease inhibitor cocktail containing 0 . 5 mM PMSF , 10 µg/ml Aprotinin , 0 . 7 µg/ml Pepstatin A , and 0 . 7 µg/ml Leupeptin . The homogenates were centrifuged 10 min at 9 , 500 g and 1 mg/ml heparin , 0 . 5% Na deoxycholate , and 0 . 5% Triton ×100 were added to the supernatant . 50 mg of lysate were deposited on a 36 ml 7% to 47% sucrose gradient in a buffer containing 20 mM HEPES ( pH 7 . 6 ) , 100 mM KCl , 5 mM MgCl2 , and 1 mM DTT . After 4 h 30 min of centrifugation at 130 , 000 g and 4°C , the gradient was divided in fractions of approximately 1 ml with a peristaltic pump . Optic density of the fractions at 260 nm was measured to establish the polysomal profile in the gradient . Fractions were finally pooled in ten fractions . An example of polysome profile is given on Figure S18 . RNAs were then extracted according to the protocol described by Clancy et al . [70] that we slightly modified . Briefly , fractions were precipitated by the addition of three volumes of ethanol and kept overnight at −80°C . After 30 min of centrifugation at 5 , 200 g , RNAs were extracted from the non-soluble fraction by classical protocol [71] . Liver RNAs were extracted and analysed by real-time quantitative RT-PCR , mostly as previously described [25] . Briefly , 0 . 5 µg of liver RNA was reverse transcribed using random hexamers and SuperScript II reverse transcriptase ( Life Technologies ) . The cDNAs equivalent to 20 ng of RNA were PCR amplified in triplicate in an ABI PRISM 7700 Sequence Detection System ( Applied Biosystem ) using the TaqMan or the SYBR Green technologies . References and sequences of the probes are given in Tables S9 and S10 , respectively . Gapdh mRNA ( total RNA ) or 28S rRNA ( polysomal RNA ) were used as controls . Liver polysomal and total RNAs were extracted independently from two mice sacrificed every 2 h during 48 h . For polysomal RNAs , we pooled fractions 1 and 2 from the ten fractions obtained during the extraction and containing heavy polysomes . 3 µg of polysomal and total RNAs from each animal from each time point were pooled . These 6 µg of polysomal and total RNAs were used for the synthesis of biotinylated cRNAs according to Affymetrix protocol , and hybridized to mouse Affymetrix Mouse Genome 430 2 . 0 arrays . The chips were washed and scanned , and the fluorescence signal analysed with Affymetrix software . Data are deposited on the Gene Expression Omnibus database under the reference GSE33726 ( http://www . ncbi . nlm . nih . gov/geo/query/acc . cgi ? token=rpwvtoqogkamwrm&acc=GSE33726 ) . The raw data of all 48 arrays were normalized together using the robust multiarray average ( RMA ) method [72] . For the analysis , we filtered out all probesets corresponding to introns using the Ensembl annotation and then only kept genes with a sufficient expression level ( we kept genes whose probe signal in the total fraction was above 5 in log2 scale ) . For the identification of circadian probesets , the 24-h Fourier component ( F24 ) and the phase were computed using established methods [73] . The associated p-value ( p ) was calculated using the Fisher test ( p = ( 1−s ) 10 ) [73] . For the identification of rhythmically translated genes , the difference between polysomal and total RNAs was subjected to Fourier analysis and we selected probesets giving a p-value inferior to 0 . 001 . In addition , we requested that the peak to trough amplitude in the polysomal signal be above 1 . 2-fold . Nuclear and cytoplasmic proteins were extracted mostly as described [25] . Briefly , liver were homogenized in sucrose homogenization buffer containing 2 . 2 M sucrose , 15 mM KCl , 2 mM EDTA , 10 mM HEPES ( pH 7 . 6 ) , 0 . 15 mM spermin , 0 . 5 mM spermidin , 1 mM DTT , and the same protease inhibitor cocktail as for polysomes extraction . Lysates were deposited on a sucrose cushion containing 2 . 05 M sucrose , 10% glycerol , 15 mM KCl , 2 mM EDTA , 10 mM HEPES ( pH 7 . 6 ) , 0 . 15 mM spermin , 0 . 5 mM spermidin , 1 mM DTT , and a protease inhibitor cocktail . Tubes were centrifuged during 45 min at 105 , 000 g at 4°C . After ultra-centrifugation , supernatants containing soluble cytoplasmic proteins were harvested , homogenised , and centrifuged for 2 h at 200 , 000 g to remove ribosomes . These supernatants constitute cytoplasmic extracts . The nucleus pellets were suspended in a nucleus lysis buffer composed of 10 mM HEPES ( pH 7 . 6 ) , 100 mM KCl , 0 . 1 mM EDTA , 10% Glycerol , 0 . 15 mM spermine , 0 . 5 mM spermidine , 0 . 1 mM NaF , 0 . 1 mM sodium orthovanadate , 0 . 1 mM ZnSO4 , 1 mM DTT , and the previously described protease inhibitor cocktail . Nuclear extracts were obtained by the addition of an equal volume of NUN buffer composed of 2 M urea , 2% nonidet P-40 , 600 mM NaCl , 50 mM HEPES ( pH 7 . 6 ) , 1 mM DTT , and a cocktail of protease inhibitor , and incubation 20 min on ice . After centrifugation during 10 min at 21 , 000 g , the supernatants were harvested and constitute nuclear extracts . 25 µg of nuclear or 12 . 5 µg cytoplasmic extracts were used for western blotting . After migration , proteins were transferred to PVDF membranes and Western blotting was realized according to standard procedures . References for the antibodies are given in Table S11 . Organs were homogenized in lysis buffer containing 20 mM HEPES ( pH 7 . 6 ) , 100 mM KCl , 0 . 1 mM EDTA , 1 mM NaF , 1 mM sodium orthovanadate , 1% Triton X-100 , 0 . 5% Nonidet P-40 , 0 . 15 mM spermin , 0 . 5 mM spermidin , 1 mM DTT , and a protease inhibitor cocktail . After incubation 30 min on ice , extracts were centrifuged 10 min at 21 , 000 g and the supernatants were harvested to obtain total extracts . 65 µg of extract was used for Western blotting . After migration , proteins were transferred to PVDF membranes and Western blotting was realized according to standard procedures . References for the antibodies are given in Table S11 . 7-methyl GTP sepharose 4B beads ( GE Healthcare ) were washed twice in the previously described liver lysis buffer . 250 µg of liver protein extracts were diluted in 500 µl of lysis buffer containing 1 mM DTT and a cocktail of protease inhibitor and incubated for 2 h on a rotating wheel at 4°C with 20 µl of beads . After incubation , cap-binding-proteins coated beads were washed five times in 500 µl of liver lysis buffer containing 0 . 5 mM PMSF and 1 mM DTT . 7-methyl GTP bound proteins were eluted by SDS-PAGE loading buffer , separated by SDS-PAGE , transferred to PVDF membranes , and analysed by Western blotting as described . Mean and standard error of the mean were computed for each time point . The rhythmic characteristics of the expression of each gene or protein were assessed by a Cosinor analysis [74] . This method characterizes a rhythm by the parameters of the fitted cosine function best approximating the data . A period of 24 h was a priori considered . The rhythm characteristics estimated by this linear least squares method include the mesor ( rhythm-adjusted mean ) , the double amplitude ( difference between minimum and maximum of fitted cosine function ) , and the acrophase ( time of maximum in fitted cosine function ) . A rhythm was detected if the null hypothesis was rejected with p<0 . 05 . In such a case , the 95% confidence limits of each parameter were computed . The Cosinor 2 . 3 software used in this study has been elaborated by the Circadian Rhythm Laboratory at University of South Carolina and is freely available at this address: http://www . circadian . org/softwar . html . The statistical significance of differences in the mesor was evaluated by a Student's t-test .
|
Most living organisms on earth present biological rhythms that play a fundamental role in the coordination of their physiology and behavior . The discovery of the molecular circadian clock gives important insight into the mechanisms involved in the generation of these rhythms . Indeed , this molecular clock orchestrates the rhythmic transcription of clock-controlled genes involved in different aspects of metabolism , for example lipid , carbohydrate , and xenobiotic metabolisms in the liver . However , we show here that the circadian clock could also exert its function through the coordination of mRNA translation . Namely , the circadian clock influences the temporal translation of a subset of mRNAs by controlling the expression and activation of translation initiation factors , as well as the clock-dependent rhythmic activation of signaling pathways involved in their regulation . These rhythmically translated mRNAs are mainly involved in ribosome biogenesis , an energy consuming process , which has to be gated to a period when the cell resources are less limited . Moreover , the role of the circadian oscillator in this process is highlighted by its direct regulation of the transcription of ribosomal protein mRNAs and ribosomal RNAs . Thus our findings suggest that the circadian clock exerts a major role in coordinating transcription and translation steps underlying ribosome biogenesis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Material",
"and",
"Methods"
] |
[
"systems",
"biology",
"biochemistry",
"biology",
"genomics",
"molecular",
"cell",
"biology",
"genetics",
"and",
"genomics"
] |
2013
|
The Circadian Clock Coordinates Ribosome Biogenesis
|
The combination of deworming and improved sanitation or hygiene may result in greater reductions in soil-transmitted helminth ( STH ) infection than any single intervention on its own . We measured STH prevalence in rural Bangladesh and assessed potential interactions among deworming , hygienic latrines , and household finished floors . We conducted a cross-sectional survey ( n = 1 , 630 ) in 100 villages in rural Bangladesh to measure three exposures: self-reported deworming consumption in the past 6 months , access to a hygienic latrine , and household flooring material . We collected stool samples from children 1–4 years , 5–12 years , and women 15–49 years . We performed mini-FLOTAC on preserved stool samples to detect Ascaris lumbricoides , Enterobius vermicularis , hookworm , and Trichuris trichiura ova . Approximately one-third ( 32% ) of all individuals and 40% of school-aged children had an STH infection . Less than 2% of the sample had moderate/heavy intensity infections . Deworming was associated with lower Ascaris prevalence ( adjusted prevalence ratio ( PR ) = 0 . 53; 95% CI 0 . 40 , 0 . 71 ) , but there was no significant association with hookworm ( PR = 0 . 93 , 95% CI 0 . 60 , 1 . 44 ) or Trichuris ( PR = 0 . 90 , 95% CI 0 . 74 , 1 . 08 ) . PRs for hygienic latrine access were 0 . 91 ( 95% CI 0 . 67 , 1 . 24 ) , 0 . 73 ( 95% CI 0 . 43 , 1 . 24 ) , and 1 . 03 ( 95% CI 0 . 84 , 1 . 27 ) for Ascaris , hookworm , and Trichuris , respectively . Finished floors were associated with lower Ascaris prevalence ( PR = 0 . 56 , 95% CI 0 . 32 , 0 . 97 ) but not associated with hookworm ( PR = 0 . 48 95% CI 0 . 16 , 1 . 45 ) or Trichuris ( PR = 0 . 98 , 95% CI 0 . 72 , 1 . 33 ) . Across helminths and combinations of exposures , adjusted prevalence ratios for joint exposures were consistently more protective than those for individual exposures . We found moderate STH prevalence in rural Bangladesh among children and women of childbearing age . This study is one of the first to examine independent and combined associations with deworming , sanitation , and hygiene . Our results suggest that coupling deworming with sanitation and flooring interventions may yield more sustained reductions in STH prevalence .
The World Health Organization recommends mass drug administration of albendazole or mebendazole to school-aged children to control soil-transmitted helminths in endemic countries; in addition , they recommend improved sanitation and hygiene to ensure long-term sustainability of deworming efforts [1] . While anthelminthics are highly efficacious in the short term , it is estimated that within six months , on average , 68% of those treated become reinfected with Ascaris , 67% with Trichuris , and 55% with hookworm [2] . Improvements to sanitation [3–8] and installation of finished flooring in homes may contribute to more sustainable reductions in STH transmission and prevalence either when delivered alone or in combination with deworming . However , to date , STH control has largely been a separate enterprise from the control of enteric pathogens through sanitation and hygiene interventions [9] . There has been a call to consider the joint effects of anthelminthic treatment and water , sanitation , and hygiene interventions [7 , 9] , yet few studies have done so rigorously [10–13] . Considering the increased concerns about the potential for resistance to anthelminthic drugs [14 , 15] , it is increasingly important to identify interventions that may more sustainably reduce transmission and minimize the duration of mass drug administration campaigns . Evidence that the combination of deworming and improved household environmental conditions results in greater risk reductions than either alone ( i . e . , evidence of a synergistic interaction ) would motivate the delivery of combined interventions to more sustainably reduce the incidence and transmission of STH . Furthermore , such evidence could inform the targeting of deworming interventions to households with environmental conditions in which deworming is most effective . Several studies have suggested that improved sanitation can reduce the risk of STH infection by reducing the shedding of helminth ova into the environment , leading to reduced transmission [7 , 16–18] . Another potential intervention that may reduce transmission is the provision of finished floors ( e . g . , cement or wood floors ) . Since STH eggs must be deposited in the soil to reach their infective stages , providing finished floors to households to replace earthen floors may reduce transmission by reducing the number of infective stages in the living space . The few studies that have systematically explored whether finished flooring is associated with STH infection found evidence of reduced risk but did not adjust for potentially strong confounders , such as household wealth [19–22] . To date , no studies have formally explored interactions between water , sanitation , and flooring interventions and deworming [23] . In Bangladesh , the Ministry of Health and Family Welfare has implemented mass drug administration of mebendazole to control STH infection in schools in 27 out of 64 districts bi-annually since 2008 . In addition , the Bangladesh Expanded Program on Immunization offers mebendazole to pre-school children in Bangladesh . To control lymphatic filariasis infection , the Ministry has offered albendazole and diethylcarbamazine in endemic areas each year to all individuals over 1 year of age who are not pregnant since 2001; as of 2008 , the program has operated in 20 districts . Prior to mass drug administration , an estimated 80% of Bangladeshi school-age children in rural areas were infected with STH [24] . To our knowledge , there have not been any systematic surveys of STH prevalence since mass drug administration began in Bangladesh . Our objectives were to estimate the prevalence of STH infection among children and women of childbearing age in rural Bangladesh and to estimate the separate and combined associations of deworming , hygienic latrines , and finished floors with STH prevalence .
In this study , we collected stool samples from a subset of individuals who participated in an ongoing cross-sectional study led by the International Centre for Diarrhoeal Disease Research , Bangladesh ( ICDDR , B ) in rural Bangladesh and conducted a secondary analysis of survey data from that study . The original study evaluated the Sanitation Hygiene Education and Water Supply in Bangladesh ( SHEWA-B ) program , which was implemented by UNICEF and the Government of Bangladesh from 2007–2012 . Local hygiene promoters visited mothers of children under five years old in underserved areas of rural Bangladesh and delivered key messages about safe water , sanitation , and hygiene practices . The program did not offer deworming or improvements to household flooring . In a small subset of villages with high poverty levels , the program offered subsidized latrines [25] . The data used in this cross-sectional study were collected in 2012 in 68 sub-districts , 19 districts , and 50 intervention and 50 control village clusters in rural Bangladesh to evaluate SHEWA-B in 2012 . The intervention , selection of control areas , eligibility , and sampling of clusters have been described elsewhere [26] . In this study , we did not stratify by whether a respondent participated in SHEWA-B or lived in a control cluster because the impact evaluation of SHEWA-B found that there was no increase in access to improved latrines , safe disposal of feces , availability of a handwashing station , or safe drinking water storage among SHEWA-B participants [25] . Thus , we considered it unlikely that SHEWA-B participation would be a potential confounder of the association between our exposures of interest and STH infection . The field team collected stool samples and questionnaires about socio-demographic information and anthelminthic treatment in October 2012 . In December 2012 , they administered a questionnaire to the same households to measure access to hygienic latrines , finished floors , and other environmental exposures . Latrine and flooring status were ascertained following stool sample collection due to field logistics constraints and the need to complete stool collection prior to national mass drug administration in early November 2012 . In each cluster , stool was collected from individuals in 17 randomly selected households; this number was determined by the sample size calculations from the original SHEWA-B evaluation . Field workers aimed to stratify sample collection by age such that six people within each of the following age and sex categories provided stool in each cluster: children 1–4 years , children 5–14 years , and women 15–49 years . We stratified by these groupings because in Bangladesh pre-school children ( 1–4 years ) are offered deworming through the Expanded Program on Immunization , and school-aged children ( 5–14 years ) are offered deworming through a separate program administered by the Ministry of Health . We collected stool from women of childbearing age because our exposure assessment focused on household-level exposures , and in rural Bangladesh , women in this age group spend most of their time at home , so these exposures are more relevant for women than men . In addition , women of childbearing age are also more likely to be exposed to STH ova shed by young children than older women . Field workers provided households with plastic sheets and stool collection tubes and returned within 24 hours to collect samples . They stored 1g of stool in 20 ml of 4% sodium acetate-acetic acid-formalin . The maximum time between defecation and stool processing was 12 hours . Samples were transported to Dhaka , Bangladesh , for laboratory analysis at the International Centre for Diarrhoeal Disease Research , Bangladesh . Helminth ova were detected using mini-FLOTAC , a copromicroscopic diagnostic technique appropriate for preserved stool [27 , 28] . Laboratory staff centrifuged samples at 1500 RPM for 3 minutes and then discarded the supernatant and suspended the sedimented stool in 20 ml of flotation solution 2 ( saturated sodium chloride ) . They then mixed the contents thoroughly and filled each of the two chambers of the mini-FLOTAC device with 1 ml of the mixed sample . Staff recorded the number of eggs of Ascaris lumbricoides , hookworm , Trichuris trichiura , and Enterobius vermicularis in each chamber . For each helminth , we averaged the number of eggs in each chamber and multiplied the number by a factor of 10 to quantify the number of eggs per gram of stool . Laboratory analyses were conducted within 9 months of stool sample collection . Outcomes included the presence of any helminth ova as well as the intensity of helminth infection . Moderate/high intensity infections were defined as ≥5 , 000 eggs/gram for Ascaris , ≥1 , 000 eggs/gram for Trichuris , and ≥2 , 000 eggs/gram for hookworm [29] . Exposures included access to a hygienic latrine , household flooring material ( earth/bamboo or cement/wood ) , and self-reported deworming in the last six months . We defined hygienic latrines as flush latrines connected to a piped sewer system , septic tank , off-set pit latrine , pit latrine with slab and functional water seal , pit latrine with slab , lid and no water seal , or a composting latrine . We defined “unhygienic” latrines as those that would likely fail to separate feces from the environment effectively including flush latrines connected to canal or ditch , pit latrines with or without a slab , no or broken water seal or a hanging latrine . This definition was developed by the International Centre for Diarrhoeal Disease Research , Bangladesh ( ICDDR , B ) and is intended to more accurately categorize latrines that isolate feces from the environment in the Bangladeshi context than the commonly used World Health Organization Joint Monitoring Programme ( JMP ) definition [30] ( see S1 Table ) . Respondents reported whether each person who provided a stool sample took deworming medication in the last six months . If so , they were asked approximately how many weeks or months ago they took deworming . Field workers recorded whether deworming was received as part of a campaign and the source of deworming ( e . g . clinic , school ) . We calculated the cluster-level deworming coverage as the percentage of respondents in the sample who reported taking deworming in the prior six months in a given cluster . To estimate cluster-level sanitation and finished floor coverage , we calculated the percentage of respondents with a hygienic latrine , finished household floor , or who reported deworming consumption in a cluster . We controlled for the following potential confounders in statistical models: age , sex , household wealth , cluster-level wealth , mother's education level , and district of residence . We used information about household assets ( e . g . refrigerator , mobile phone ) to develop an index of household wealth using principal components analysis [31] ( see S2 Table ) . Households in the lowest three quintiles of the first principal component were classified as lower household wealth and those in the highest two quintiles were classified as higher household wealth . Cluster-level wealth was calculated as the percentage of households in the two highest quintiles of household wealth . Since age- and sex-specific STH prevalence estimates were not available for the study areas , we assumed the prevalence of all helminths to be 50% . We assumed a design effect of 2 . 6 based on intraclass correlation coefficients estimated in a study measuring STH infection in a similar population in India [32] since such information for Bangladesh was not available . Because our study was nested within the ongoing SHEWA-B evaluation , our calculations assumed a fixed sample size of 1 , 700 ( 100 village clusters x 17 individuals per cluster ) . Under these assumptions , the precision associated with an estimate of prevalence of 50% is ±4% . We calculated pooled and age- and sex-specific prevalence by species of helminth . To examine the association between prevalence and cluster-level variables , we produced scatter plots of the observed variables and used locally weighted scatter plot smoothing ( LOWESS ) with normal-based pointwise 95% confidence bands to explore patterns in each scatter plot [33] . We also estimated the intraclass correlation coefficient for each STH infection within each cluster using a one-way analysis of variance . To estimate adjusted prevalence ratios we used modified Poisson regression [34] adjusting for the potential confounders defined above . For each of the three exposures , we adjusted for the other two exposures in each model ( e . g . , the models for deworming were adjusted for hygienic latrines and finished floors ) . We estimated robust standard errors clustered at the village level to account for potential within-village outcome correlation . We excluded individuals with missing outcomes from the analysis , which assumes that they were missing at random conditional on covariates in our model . Standard statistical models for binary outcomes predict outcomes and assess statistical interaction on the multiplicative scale . In this study , we chose to estimate interaction on the additive scale , which is useful when one is interested in the extent to which a primary exposure may yield greater health improvements by introducing a secondary exposure . In our study population , we consider deworming to be a primary exposure due to ongoing school-based deworming activities in Bangladesh . Our approach allows us to assess whether secondary exposures in addition to deworming , such as access to hygienic latrines , were associated with lower STH prevalence than deworming alone [35] . Specifically , we estimated the relative excess risk due to interaction ( RERI ) , a measure of additive interaction [36] . When the exposures of interest are associated with only a lower or higher prevalence , an RERI>0 indicates a synergistic interaction between exposures [37 , 38] . If exposures could be associated with either increased or decreased prevalence , then the RERI must be greater than 1 for synergistic interaction to be present [37 , 38] . Since we expected associations to be protective , we recoded variables prior to RERI calculation so that the stratum with the prevalence ratio furthest from the null was reassigned as the reference group [39] . Because data were clustered at the village level , we used the bootstrap and resampled clusters to estimate 95% confidence intervals [40 , 41] . We did not estimate confidence intervals for any point estimates for which there were strata with fewer than 5 units . Analyses were conducted in Stata version 12 and in R version 3 . 1 . 3 . This study was approved by the Institutional Review Board ( IRB ) at the Centers for Disease Control and Prevention ( CDC ) ( Protocol #6061 ) . Study participants provided written consent . The CDC IRB approved this consent procedure . We offered a single dose of mebendazole to all participants who provided a stool sample .
The field team collected stool samples and demographic information from 1 , 795 individuals in October 2012 and collected exposure information from 1 , 655 individuals in December 2012 ( Fig 1 ) . There were 1 , 630 individuals in the complete dataset; 140 individuals were not home during follow-up , and 25 identification numbers were mismatched between survey rounds . The number of missing observations for each variable of interest is listed in S3 Table . Less than half ( 40% , n = 656 ) of mothers of the youngest child in each household had received at least a primary education . About a third of households ( 32% , n = 527 ) had access to a hygienic latrine , and 13% ( n = 207 ) of households had finished floors . Respondents reported that 49% of children 1–4 years , 52% of children 5–14 years , and 21% of women of childbearing age took deworming medication in the prior six months . Child caregivers reported that less than half ( 47% ) of school-age children took deworming drugs at school . Approximately one-third ( 32% ) of individuals sampled had an STH infection , and 9% had multiple infections ( Table 1 ) . Across all age groups , Trichuris was most prevalent , infecting 17% of children 1–4 years , 28% of children 5–14 years , and 18% of women of childbearing age . For all helminths and age groups , fewer than 2% had moderate/heavy intensity Ascaris or Trichuris infections; there were no moderate/heavy intensity hookworm infections . Prevalence of Ascaris and Trichuris were highest in areas in Dhaka and northern Chittagong divisions ( Fig 2 ) . We estimated associations between individual exposures and STH infection ( Table 2 ) . Deworming was associated with 47% lower Ascaris prevalence ( 95% CI 29% , 60% ) ; adjusted associations with hookworm and Trichuris were null . There were no statistically significant associations with hygienic latrine access for any helminth . Finished floors were associated with 44% lower Ascaris prevalence ( 95% CI 3% , 68% ) . There was no statistically significant association between finished floors and hookworm or Trichuris . Participation in SHEWA-B was not associated with STH infection ( S4 Table ) . To explore potential interactions among deworming , hygienic latrines , and finished floors , we estimated adjusted prevalence ratios for each separately and in combination and the relative excess risk due to interaction RERI ( Tables 3 and 4 ) . The combination of deworming and hygienic latrines was associated with lower Ascaris and Trichuris prevalence than associations with separate exposures , although this association was only statistically significant for Ascaris ( Table 3 , Panel A; Fig 3 ) . For example , deworming without hygienic latrine access ( denoted D+L- in Fig 3 ) was associated with 45% lower Ascaris prevalence ( 95% CI 23% , 60% ) . Ascaris prevalence was equivalent for those with hygienic latrine access who did not take deworming ( D-L+ ) . The combination of deworming and hygienic latrine access ( D+L+ ) was associated with 59% lower Ascaris prevalence ( 95% CI 27% , 76% ) . While individual and combined exposures were associated with lower hookworm prevalence , the associations did not follow the same pattern and were not statistically significant . Our estimates of the RERI for Ascaris and Trichuris were less than zero and not statistically significant , indicating no clear evidence of synergistic interaction between deworming and hygienic latrines on the additive scale . However , there was some suggestion of synergistic interaction for hookworm ( RERI = 0 . 49; 95%CI -0 . 73 , 1 . 16 ) , although the confidence interval included 0 . For the combination of deworming and finished floors , we found the same pattern for all three helminths: the combination was associated with a lower prevalence than either exposure on its own ( Table 3 , Panel B; Fig 3 ) . Deworming without finished floors ( denoted D+F- in Fig 3 ) was associated with 47% lower Ascaris prevalence ( 95%CI 29% , 61% ) . Finished floors without deworming ( D-F+ ) was associated with 40% lower Ascaris prevalence ( 95%CI -15% , 68% ) . The combination of deworming and finished floors ( D+F+ ) was associated with 72% lower Ascaris prevalence ( 95%CI 89% , 25% ) . For Ascaris , the RERI was 0 . 56 ( 95%CI -3 . 64 , 2 . 40 ) . This pattern was also evident for hookworm and Trichuris . We plotted the cluster-level prevalence of each helminth across the observed range of cluster-level deworming , hygienic latrine , and finished floor coverage ( Figs 2 and S1 and S2 ) . As hygienic latrine coverage increased , Ascaris and hookworm prevalence remained approximately the same and Trichuris prevalence increased slightly ( Fig 4 ) . As deworming and finished floor coverage increased , there was no substantial change in the prevalence of any helminth ( S1 and S2 Figs ) . The village cluster level intraclass correlation coefficients were 0 . 11 ( 95% CI 0 . 08 , 0 . 16 ) for Ascaris , 0 . 02 ( 95% CI 0 . 00 , 0 . 05 ) for hookworm , and 0 . 21 ( 95% CI 0 . 15 , 0 . 27 ) for Trichuris .
The prevalence of any STH among school-aged children was 40% compared to 80% in rural areas reported by the Ministry of Health and Family Welfare prior to the initiation of school-based deworming [24] . The prevalence we observed is consistent with studies of school-based deworming with one-year follow-up and very high coverage [42–44] . Transmission theory [45–47] and empirical findings [48 , 49] suggest that prevalence decreases in pre-school age and adult populations when deworming coverage is high . We found that 26% of children 1–4 years and 30% of women of childbearing age had an STH infection . The similar prevalence in these two groups to the prevalence among school-aged children may suggest that , despite low levels of intensity , transmission is still ongoing in those population subgroups . We observed associations between STH infection and deworming , finished floors , and hygienic latrines that were generally consistent with existing studies . Contrary to what we would expect based on modeling studies [45–47] , we did not find evidence of an association between STH prevalence and coverage of exposures at the cluster level . We expected that prevalence would decrease as the coverage of deworming , hygienic latrines , and finished floors increased . Such a pattern could be explained by herd effects of these exposures , which result from decreased shedding of infective stages in feces into the environment and reduced transmission [47] . Interestingly , our estimates of intraclass correlation coefficients suggest that cluster membership accounts for 25% of Trichuris prevalence and 11% of Ascaris prevalence; these findings are similar to the intraclass correlation coefficients reported in the literature [60] . Thus , village membership appears to be an important predictor of STH infection despite the lack of associations with cluster-level exposure coverage . It is possible that we did not find cluster-level associations because herd effects were small or absent for these exposures . It is also possible that such herd effects might have been higher if STH prevalence were higher . Alternatively , because we only collected data for 16 households per cluster on average , our sample might not have been sufficient to characterize cluster-level prevalence patterns . Additionally , variation in local population density in our sample could explain the lack of association because herd effects are likely to be stronger in densely populated areas . We did not account for population density in this analysis , and further work is needed to explore its role as a potential effect modifier . This study is subject to several limitations . First , deworming consumption was self-reported over a six-month period . There is evidence that recall of medication consumption is under-reported with longer recall periods [61] . It is also possible that reporting was subject to courtesy bias so that deworming was over-reported . Given that it is unlikely that respondents knew if they had STH infection at the time of deworming reporting , we posit that it is unlikely that misclassification of deworming consumption differed by STH infection status . If non-differential misclassification occurred , it would bias point estimates towards the null [62] . It is also possible that individuals who were dewormed in the prior six months were reinfected prior to stool collection in this study . For this reason and because of the possible misclassification of deworming due to poor recall , the associations between deworming and STH infection do not necessarily measure the reduction in STH attributable directly to deworming . Second , on average , stool samples were stored for 6 months prior to mini-FLOTAC analysis ( min = 4 , max = 8 months ) . This long storage period may have resulted in degradation of ova and underestimated prevalence , particularly for hookworm [63] . The sensitivity of mini-FLOTAC is estimated to be 75 . 5% for Ascaris , 79 . 2% for hookworm , and 76 . 2% for Trichuris [64] . Thus , given the sensitivity of the assay and the long storage period prior to analysis , our prevalence estimates are likely a lower bound of the true values . In addition , the sample size was powered to estimate prevalence but not to estimate interactions between exposures . Thus , our estimates of the RERI were in most cases underpowered , particularly for improved floors , which were relatively rare in this population . Third , exposure measurement occurred following outcome measurement . This is suboptimal because it is possible that the outcome status of an individual would trigger a change in exposure status , leading to reverse causation . However , for the exposures measured—access to a hygienic latrine and finished floors—we consider it highly unlikely that the respondents' exposure status changed in the two months following outcome measurement . Furthermore , respondents were likely unaware of their outcome status throughout the study , so it is unlikely that they changed their sanitation infrastructure or flooring material because of their outcome status . Fourth , our analysis assumes that the sanitation and flooring infrastructure we observed were present within the past six months ( the recall window for deworming consumption ) . We consider this to be a reasonable assumption; however , if these exposures were misclassified , prevalence ratios would be biased towards the null and the effects on measures of interaction would be unpredictable [65] . This study provides the first estimates of STH prevalence following the initiation of mass drug administration for STH control among children and women of childbearing age in representative rural areas of Bangladesh . Our results suggest that combining deworming with sanitation and flooring interventions may yield greater reductions in STH prevalence than deworming alone .
|
Soil-transmitted helminth infections remain prevalent in many low-resource areas of the world . The World Health Organization recommends that schoolchildren in countries where these infections remain common receive deworming medication two times a year . However , previous research has shown that people who live in countries where these infections are common are frequently reinfected within 6 months of taking deworming medication . Programs that improve sanitation and hygiene might help complement deworming programs to reduce reinfection and prevent transmission . We conducted a survey of women and children in rural Bangladesh to understand potential sanitation and hygiene interventions that could complement deworming . We found that people who took deworming medication and had access to a hygienic latrine had a lower worm infection prevalence than people who only took deworming medication . We also found that people who took deworming medication and had a house with a finished floor had a lower prevalence than people who only took deworming medication . Our results suggest that coupling deworming with sanitation and flooring interventions may be a more successful strategy for reducing STH transmission in the long run .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
The Interaction of Deworming, Improved Sanitation, and Household Flooring with Soil-Transmitted Helminth Infection in Rural Bangladesh
|
Human T cell lymphotropic virus-1 ( HTLV-1 ) primarily infects CD4+ T cells , causing inflammatory disorders or a T cell malignancy in 5% to 10% of carriers . The cytotoxic T lymphocyte ( CTL ) response is a key factor that controls the viral load and thus the risk of disease . The ability to detect the viral protein Tax in primary cells has made it possible to estimate the rate at which Tax-expressing infected cells are eliminated by CTLs in persistently infected people . However , most HTLV-1-infected cells are Tax–at a given time , and their immunophenotype is poorly defined . Here , we aimed to identify a cell-surface molecule expressed by both Tax+ and Tax–HTLV-1-infected cells and use it to analyse the CTL response in fresh peripheral blood mononuclear cells . Cell adhesion molecule 1 ( CADM1/TSLC1 ) was the best single marker of HTLV-1 infection , identifying HTLV-1-infected cells with greater sensitivity and specificity than CD25 , CCR4 or ICAM-1 . CADM1+CD4+ T cells carried a median of 65% of proviral copies in peripheral blood . In a cohort of 23 individuals , we quantified the rate of CTL-mediated killing of Tax+ and Tax−CADM1+ cells . We show that CADM1 expression is associated with enhanced susceptibility of infected cells to CTL lysis: despite the immunodominance of Tax in the CTL response , Tax+CADM1– cells were inefficiently lysed by CTLs . Upregulation of the CADM1 ligand CRTAM on CD8+ T cells correlated with efficient lysis of infected cells . Tax–CADM1+ cells were lysed at a very low rate by autologous CTLs , however , were efficiently killed when loaded with exogenous peptide antigen . High expression of CADM1 on most HTLV-1-infected cells in the face of enhanced CTL counterselection implies that CADM1 confers a strong benefit on the virus .
Human T-lymphotropic virus 1 ( HTLV-1 ) is a retrovirus that predominantly infects CD4+ T cells . An estimated 10–20 million people are infected , with regions of high prevalence including Japan , Africa , the Caribbean and South America . The viral burden ( proviral load , PVL ) is strongly correlated with the risk of disease [1] . Between 1% and 6% of HTLV-1-infected individuals develop a T cell malignancy known as adult T cell leukemia/lymphoma ( ATL ) , and an additional 2–3% suffer from a variety of inflammatory disorders , the most prevalent of which is HTLV-1 associated myelopathy/tropical spastic paraparesis ( HAM/TSP ) . Although HTLV-1 was the first retrovirus observed to be pathogenic in humans , both effective treatment and a vaccine remain elusive . HTLV-1 persists within an infected individual by infectious spread across the virological synapse and by mitotic replication of infected cells [2 , 3]; virus particles are usually undetectable in peripheral blood [4] . The proviral integration site imparts each infected T cell clone with a different pattern and intensity of viral gene expression [5 , 6] . Of these , Tax and HTLV-1 basic leucine zipper protein ( HBZ ) , two regulatory proteins , play an important role in viral pathogenesis . The transcriptional transactivator , Tax , encoded in the positive strand in the regulatory ( pX ) region of the virus , controls the expression of viral proteins ( Pol , Gag and Env ) as well as many host genes [7] . The negative strand-encoded accessory gene HBZ can inhibit Tax function and modify transcription of various host genes [7] . The PVL of HTLV-1 reaches a stable ( ‘set-point’ ) level in each individual [8] , which is maintained by the equilibrium between the proliferation of infected cells and their elimination by activated cytotoxic T lymphocytes ( CTLs ) [9 , 10] . Tax , which is immunodominant , is subject to strong selection pressure from the autologous CTL response [11] , and a high lytic efficiency of HTLV-1-specific CTLs ( defined as the rate of clearance of Tax+CD4+ T cells/% CD8+ T cells/day ) is associated with low PVL and a low risk of HAM/TSP [12] . Tax expression in fresh peripheral blood mononuclear cells ( PBMCs ) is typically low in asymptomatic HTLV-1 carriers ( ACs ) and is silenced in ~50% of ATL clones [13–15] . In addition , CTL-selected Tax sequence variants are generally defective in their transactivating function [11] , impairing expression of positive strand-encoded viral genes . In contrast , HBZ is persistently expressed at low levels under the control of SP1 transcription factors [16] . HBZ minimizes its exposure to the host immune response by virtue of low protein expression , low immunogenicity and poor binding to Class 1 MHC alleles [17 , 18] . Compared with Tax-specific CTLs , HBZ-specific CTLs are present at lower frequency in the peripheral blood and kill fewer HTLV-1-infected cells in vitro [17 , 18] . However , HBZ-specific CTLs appear to be more effective in controlling HTLV-1 during chronic infection in vivo [18 , 19] . A limitation of the previously described assay of anti-HTLV-1 CTL lytic efficiency [12] is that the identification of HTLV-1+ cells was defined by Tax expression , so the lysis rate of Tax−cells–including those that express the key CTL target antigen HBZ–was not quantified . To date , there has been no practicable means to differentiate an HTLV-1-infected cell from an uninfected cell without destroying it in the process , either by permeabilizing it to detect Tax or other viral proteins , or by extracting DNA to detect the viral genome . The aim of the present study was to identify the cell-surface phenotype that most efficiently distinguishes viable HTLV-1-infected cells from uninfected cells , to allow more accurate analysis of the cellular immune response to the virus and the frequency and phenotype of HTLV-1-infected cells . Several activation markers , costimulatory receptors , chemokine and interleukin receptors and adhesion molecules are strongly expressed on HTLV-1-infected cells and serve as possible markers of HTLV-1 infection . These molecules include the IL-2 receptor α chain ( CD25 ) [20] , chemokine receptor 4 ( CCR4/CD194 ) [21 , 22] and intercellular adhesion molecule 1 ( ICAM-1/CD54 ) [23 , 24] . CD25 is induced by Tax , and the CD4+CD25+ population is typically ~90% HTLV-1-infected [20 , 25] . CCR4 is not directly induced by Tax [21] , but CCR4+ cells appear to be preferentially infected by HTLV-1 , perhaps because they are attracted to infected cells as a result of Tax-induced expression of CCL22 [26] . Yamano et al . ( 2009 ) recommended the use of both CD25 and CCR4 to isolate HTLV-1-infected cells with high purity [27] . ICAM-1 , also induced by Tax , aids in the formation of the virological synapse [28–30] . However , although CD25 , CCR4 and ICAM-1 are highly expressed in HTLV-1 carriers with a high viral load , they are also expressed on activated T cells and regulatory T-cells during inflammatory and immune responses to pathogens . Cell adhesion molecule 1 ( CADM1 ) , also known as tumour suppressor in lung cancer 1 ( TSLC1 ) , nectin-like protein 2 ( Necl-2 ) , immunoglobulin superfamily member 4 ( IGSF4 ) , synaptic cell adhesion molecule ( SynCAM ) or spermatogenic immunoglobulin superfamily ( SgIGSF ) is a cell-surface molecule that has recently been proposed as a marker of malignant cells in ATL patients [31–33] . CADM1 is a nectin-like cell adhesion molecule of the immunoglobulin superfamily . It was initially identified as a tumour suppressor of a range of solid cancers ( breast , ovarian , cervical and colorectal carcinomas , melanomas and neuroblastomas ) [34] . Although CADM1 is expressed in a wide range of tissues , it is usually absent in leukocytes; typically 1% of T cells are CADM1+ [31 , 32 , 34–36] . CADM1 can form both homophilic and heterophilic interactions that stabilise cell-to-cell interactions . Outside-in CADM1 signalling can trigger the rearrangement of the actin cytoskeleton to alter cell polarity and motility [37] . CADM1 also interacts with another immunoglobulin superfamily cell-surface protein , Class-1 MHC-restricted T-cell associated molecule ( CRTAM/CD355/Cytotoxic and Regulatory T Cell Molecule ) on activated NK , CD8+ T and NKT cells . This interaction enhances NK cell-mediated cytotoxicity and IFN-γ secretion by CD8+ T cells [38 , 39] . CADM1 is consistently expressed on T-cells in ATL patients and in HTLV-1-transformed cell lines [31 , 32]; the percentage of CD4+CADM1+ cells in patients with ATL positively correlates with both the PVL and the frequency of morphologically abnormal lymphocytes [32] . Immunohistochemical staining of CADM1 in organs of NOD-SCID/γcnull mice transplanted with ATL cells showed a correlation between the level of CADM1 expression on ATL cells and the ability of the ATL cells to invade solid organs such as the liver , ovaries and lungs . It has been postulated that the homophilic adhesion promotes the growth of ATL cells in these organs , in which CADM1 expression is indigenous [31 , 40] . In the present study we evaluated the specificity of CD25 , CCR4 , ICAM-1 and CADM1 as markers of HTLV-1 infection in uncultured PBMCs of individuals with non-malignant HTLV-1 infection ( ACs and patients with HAM ) . We then used the surface markers with the best combination of sensitivity and specificity to quantify and compare the rate at which autologous CTLs lysed Tax-expressing and non-expressing HTLV-1-infected cells .
To identify a marker of HTLV-1-infected cells , we first analysed the expression of candidate cell-surface molecules ( CADM1 , CD25 , CCR4 and ICAM-1 ) on uncultured PBMCs . PBMCs from 13 ACs , 11 patients with HAM and 8 uninfected individuals were analysed by flow cytometry , and the frequency of cells expressing the above markers within total PBMCs was calculated . Patients with HAM had significantly higher frequencies of CADM1+CD4+ and CADM1+CD8+T cells in PBMCs compared to uninfected individuals ( Fig 1A ) . A positive correlation was observed between the frequency of CADM1+CD4+ T cells and the PVL , which accounts for the observation that the frequency of CADM1+CD4+ cells was greater in patients with HAM than in ACs ( Fig 1B ) . While there was no significant difference between ACs and uninfected individuals , patients with HAM had a greater frequency of CD25+CD4+ and CCR4+ CD4+ T cells than uninfected individuals ( S3A Fig ) . A similar pattern was observed in CD8+ T cells ( S4 Fig ) . The frequency of total ICAM-1+CD4+ T cells in patients with HAM was higher than in ACs but was not significantly different from the frequency in uninfected individuals . Despite these differences , there was no significant correlation between the PVL and the frequency of CD25+ , CCR4+ or ICAM-1+ cells ( S3B Fig ) . The frequency of CCR4+ CD4+ and ICAM-1+ CD4+ T cells was significantly greater than the PVL , i . e . these markers are not specific to HTLV-1 infection . PBMCs from the same individuals were cultured for 18 hours to allow spontaneous viral gene expression in the presence of concanamycin A ( CMA ) to inhibit CTL-mediated lysis . Infected cells were detected by intracellular staining of the viral protein Tax . Tax expression was most frequent in cells expressing high levels of CADM1 ( Fig 2A ) . There was no significant change in the frequency of CADM1+CD4+ T cells during incubation in vitro . We quantified the frequency of expression of CD25 , CCR4 , ICAM-1 and CADM1 in Tax+ and Tax–CD4+ T cells . These data were used to calculate the sensitivity ( % marker+Tax+/total Tax+ ) and the specificity ( % Tax+marker+/total marker+ ) of detection of Tax+ cells by each respective surface marker ( S5 Fig ) . The representative dot plots in Fig 2B show that most Tax+ cells expressed high levels of CADM1 , CD25 , CCR4 and ICAM-1 . CCR4 and ICAM-1 were expressed at high frequencies on Tax–CADM1– cells ( Fig 2B ) . Thus , CCR4 and ICAM-1 had high sensitivity but lacked the specificity required for an optimal marker . However , a high intensity of ICAM-1 expression ( ICAM-1high gate , Fig 2B ) identified Tax+ cells with high sensitivity ( 84% ) and specificity ( 77% ) . The intensity of CD25 expression was correlated with that of Tax expression ( S6 Fig ) , and CD25 was generally not expressed on CADM1– cells . However , CD25 identified only a third of the Tax−infected population detected by CADM1 ( Fig 2B ) , and the low intensity of expression of CD25 made it difficult to distinguish CD25+ cells from CD25– cells . Given that the frequency of CADM1+CD4+ T cells correlated with the proviral load , but exceeded the frequency of Tax-expressing cells , we wished to quantify the frequency of HTLV-1 infection in the CADM1+CD4+ cells . In order to do this , we flow-sorted uncultured live CD3+ T cells into the following subsets: CADM1+/–CD4+ and CADM1+/−CD8+ T cells , and quantified the number of HTLV-1 copies per cell ( Fig 3A ) . We analysed 12 HTLV-1 infected donors ( 6 AC , 6 HAM ) who had proviral loads between 3% and 31% of PBMC . Sorted CADM1+ lymphocytes had a consistently high proviral load: CADM1+CD4+ cells contained a median of 1 . 2 copies of HTLV-1 per cell , and CADM1+CD8+ cells contained a median of 1 . 3 copies ( S8A Fig ) . In contrast , CADM1–CD4+ and CADM1–CD8+ cells carried a median of 0 . 05 and 0 . 02 copies per cell respectively . As proviral expression and virion production is negligible in chronically infected donors [4]; and patient derived infected primary T cell clones carry on average one proviral copy per cell [41] , we interpret that each viral copy detected in this assay represents a single cell with one integrated provirus . Although CADM1+CD4+ and CADM1+CD8+ cells appear to be 100% infected , the frequency of CADM1+CD4+ cells was approximately tenfold higher than that of CADM1+CD8+ cells ( S8B Fig ) . We calculated that a median of 65% of the proviral load in lymphocytes was carried in CADM1+ CD4+ cells ( Fig 3A and S8 Fig ) ; this proportion was consistent across all proviral loads tested ( S8C Fig ) . In addition , an independent cell-sorting experiment gave a very similar median frequency of 64% ( n = 6 individuals , S9 Fig ) . The next largest reservoir of infected cells were CADM1–CD4+ cells ( median 22% of total load ) , which had a low per-cell proviral load ( 0 . 05 copies/cell ) , but were the most abundant population identified in this analysis ( Fig 3 and S9B Fig ) . Although CADM1+CD8+ cells are heavily infected , their contribution to the proviral load was highly variable between donors ( contributing between 1% and 37% of total proviral load , median contribution 7 . 7% ) ; however , CADM1 remains a useful marker for enriching infected CD8+ cells . If Tax positivity alone is used to identify cells carrying HTLV-1 , on average 19% of infected cells would be detected , a majority of which ( ~90% of Tax+ cells ) are CADM1+ ( Fig 3B ) . If CADM1 alone is used as a marker , we detect 65% of the infected cells , including both Tax+ and Tax–cells . Considering the facts that ( i ) Tax protein expression is usually undetectable in fresh PBMCs , ( ii ) Tax protein is expressed in only a proportion of infected cells even after in vitro culture and ( iii ) CADM1 identifies both Tax+ and Tax−infected cells with high purity even in fresh PBMCs , we conclude that CADM1 is the best single cell-surface marker for HTLV-1 infection identified to date . In order to further elucidate the role of HTLV-1 in the observed CADM1 expression , we tested whether a range of candidate HTLV-1 gene products could induce CADM1 expression on a CADM1 negative human T cell line . We observed that transfection with Tax or HBZ alone was not sufficient to induce CADM1 expression ( Fig 4 ) . As HBZ mRNA possesses biological activities distinct from HBZ protein , we also utilised a construct which expresses HBZ S1 with a mutation in the initiation codon:HBZ-TTG [42] . This construct did not induce CADM1 expression on unstimulated CEM cells . We hypothesised that HTLV-1 might modify the expression in response to an external stimulus . When stimulated with PMA/CAI , a fraction of CEM cells expressed CADM1 . Tax transfection induced a twofold increase in the number of cells expressing CADM1 ( Fig 4A ) . The intensity of CADM1 expression by Tax-transfected PMA/CAI stimulated cells was also significantly higher than the intensity of GFP transfected cells which were stimulated in the same way ( Fig 4B ) . HBZ protein , but not HBZ RNA , consistently downregulated PMA/CAI induced CADM1 ( 3/3 biological replicates ) , and could reverse Tax-mediated enhancement of CADM1 expression ( 3/3 biological replicates ) . Previous work indicates that an efficient Tax-specific CTL response is associated with a lower PVL [12 , 43] and a lower risk of HAM . However , most infected cells do not express Tax ex vivo or in short-term culture in vitro . MacNamara et al . ( 2010 ) showed that an HBZ-specific CTL response was associated with effective control of HTLV-1 in vivo [18] . We therefore wished to measure the rate at which Tax–CADM1+ cells are killed by CTLs . We used the infected cell elimination assay [12] to quantify the selective pressure exerted by autologous CD8+ T cells on the different cell subpopulations in vitro , using samples from 23 infected individuals ( 11 ACs and 12 patients with HAM ) Fig 5A and 5B . Incorporation of counting beads in the assay enabled us to quantify the cells of each different phenotype and hence calculate the rate at which each population was lysed by CTL ( Fig 5C ) . The rate of CTL lysis of Tax+CADM1+ cells was significantly higher than that of any other population identified ( %target population / % CD8+ T cell per day , Fig 5C ) , with a median of 45% of Tax+CADM1+ cells lysed over 18 hours at each individual's physiological CD4:CD8 ratio . The rate of lysis of Tax–CADM1+ and Tax+CADM1– populations was not significantly different from zero , however , the rate of lysis of Tax+CADM1+ cells was significantly correlated with that of Tax–CADM1+ cells ( Fig 5D ) in HAM patients . The mean CTL lytic efficiency in the ACs in this cohort was lower than in previous studies , which may explain the observed lack of correlation in this group . Finally , although the kinetics of viral gene expression in CADM1– cells was indistinguishable from CADM1+ cells ( S7 Fig ) , Tax+CADM1– cells , which were present at the lowest frequency , consistently increased in frequency on co-culture with CD8+ T cells ( from a median of 0 . 6% of live CD4+ cells in CD8 depleted fraction to a median of 0 . 83% of live CD4+ cells at the physiological CD4:CD8 ratio ) . Thus , although the Tax+CADM1– cells expressed the immunodominant CTL target antigen Tax , the absence of CADM1 was associated with escape from CTL lysis . As CADM1 expression on the surface of the infected cell is associated with its susceptibility to CTL-mediated lysis , we quantified the expression of CADM1 and CRTAM on the surface of CD8+ T cells from individuals infected with HTLV-1 . As CRTAM is not expressed ex vivo , and is upregulated on antigen recognition [38] we quantified CRTAM after autologous CD8+ T cells were cultured with HTLV-1-infected cells at each individuals’ physiological CD4+:CD8+ ratio for 18 hours . As previously reported by others , a majority of the CRTAM+CD8+ T cells co-expressed CADM1 ( Fig 6 ) . The frequency of CADM1+ CD8+ cells always exceeded the frequency of CRTAM+CD8+ with a median of 21% of live CD8+ expressing CADM1 and 2% CRTAM post incubation , and the frequency of CRTAM+CD8+ but not CADM1+CD8+ T cells positively correlated with the rate of lysis of the Tax+CADM1+ cells . These data indicate that after coincubation with autologous targets at the physiological CD4+CD8+ ratio , activated antigen-specific CD8+ T cells from HTLV-1-infected patients express the ligand for CADM1 . To test whether expression of CADM1 is associated with enhanced target cell killing by CTLs specific to an antigen unrelated to HTLV-1 , we pulsed primary HLA-A*0201+CD4+ T cells from HTLV-1+ donors with a HLA-A*0201-restricted CMV peptide and co-incubated them with a CMV-specific CTL clone . Even at very low concentrations of peptide , CADM1+CD4+cells were highly sensitive to CTL-mediated lysis: indeed , the Tax–CADM1+ cells were more significantly susceptible to CTL lysis than the Tax+CADM1+ cells at 0 . 0002μM and 0 . 002μM peptide ( Fig 7 ) . In contrast , the smaller Tax+CADM1– population escaped lysis and increased in frequency at lower peptide concentrations , but became increasingly susceptible to lysis at peptide concentrations greater than 0 . 002μM . These data indicate that CADM1 expression is more important than Tax-induced factors ( for example , ICAM-1 ) in determining a cell’s susceptibility to CTL-mediated lysis: CADM1–CD4+ T cells required at least 100–1000 fold more antigen presentation on the surface to elicit the same level of CTL lysis observed in the CADM1+CD4+ population . The CTL clone used in this assay was CADM1–CRTAM−in the absence of stimulation , but upregulated both CADM1 and CRTAM after anti-CD3 stimulation .
Since a typical HTLV-1-infected individual has between 104 and 105 distinct clones of infected T cells [44] , it is unlikely that a single cell-surface marker can identify all the infected cells in an individual , owing to between-clone variation in proviral expression . Nevertheless , the results presented here show that expression of the single surface marker CADM1 identifies approximately 70% of HTLV-1-infected cells ( CD4+ and CD8+ ) in the peripheral blood . Although most HTLV-1-infected cells were CCR4+ and ICAM-1+ , these markers lacked the specificity required of a single marker for infected cells: the frequency of CCR4+ and ICAM-1+ cells significantly exceeded the PVL in most HTLV-1-infected individuals . High-intensity ICAM-1 expression efficiently identifies Tax-expressing cells after incubation of PBMCs of infected individuals in vitro for 6 hours; however , ICAM-1 is not upregulated on Tax−cells . Yamano et al . proposed the use of CD25 alone or in combination with CCR4 to isolate infected cells [20 , 27] . While our data confirm that a high proportion of CD4+CD25+CCR4+ cells are infected , these cells accounted for a small proportion of the proviral load . The use of CADM1 alone in fresh cells not only identified an average of two-thirds of the cells identified by the use of CD25 and CCR4 but also detected a much larger population of infected CD25–CCR4+ cells . That is , the use of CADM1 alone identified twice the number of infected cells identified by the combination of CD25 and CCR4 . CADM1 also identified a population of CD8+ cells which were heavily infected with the virus . CADM1 has been reported as a surface marker of ATL cells [31–33] . The results of the present study extend the use of CADM1 as a marker of HTLV-1 infection to those without ATL . Kobayashi et al . reported that CADM1+ cells were frequently HTLV-1 infected and that the percentage of CADM1+ cells ( both CD7dim and CD7– ) reflected disease status [45 , 46] . These authors also suggested that the appearance of CADM1+ cells is a marker of progression to ATL . However , we detected CADM1 expression in the PBMCs of all ACs and patients with HAM tested . We hypothesise that although ATL cells may arise from CADM1+ cells ( expanded clones in ATL patients express CADM1 [31] ) , the expression of CADM1 in ACs or patients with HAM does not presage the onset of ATL . Rather , our data indicate that the frequency of CADM1+ cells reflects the PVL; a higher PVL is associated with a higher risk of ATL . It remains unknown what induces CADM1 expression in HTLV-1-infected cells . There are conflicting reports on whether Tax induces CADM1 . While we and others ( Nakahata et al . 2012 ) failed to induce CADM1 in CEM , JPX-9 , MOLT4 and 293T cells on introduction of a Tax expression vector [32] , Pujari et al . ( 2015 ) found elevated levels of CADM1 in murine embryonic fibroblasts and Jurkat T cells upon lentivirus-mediated Tax expression [47] . Although we observed that Tax+ cells are mostly CADM1+ and Tax+CADM1+ cells express on an average 2–3 fold higher intensity of CADM1 than Tax–CADM1+ cells ( S7 Fig ) , we conclude that Tax alone does not induce CADM1 in resting primary PBMCs , for the following reasons: ( i ) when PBMCs were cultured for 18 hours in vitro , the percentage of CADM1+CD4+ T cells did not increase with Tax expression; ( ii ) a significant proportion of CADM1+ cells did not express Tax; ( iii ) Tax is silenced , mutated or deleted in ~65% of ATL cases [13–15] whereas CADM1 is expressed in virtually all cases [31 , 33] . We observed that HBZ protein could downregulate CADM1 expression in response to an external stimulus , even in the presence of Tax . We postulate that the ability of the virus to modulate the level of CADM1 expression is advantageous to viral persistence . Indeed , it has recently been demonstrated that CADM1 plays an important role in the constitutive NF-κB activation observed in HTLV-1-infected cells [47] . To analyse the CD8+ T-cell response to HTLV-1-infected cells , we used a modified version of the functional assay previously described [12] . The modifications provided two significant advantages over the original assay: ( i ) the use of counting beads made it possible to enumerate cells in several target populations ( defined by different phenotypes ) , and thus to quantify the rate at which they are killed by CTLs . ( ii ) The use of CADM1 made it possible to estimate the rate of lysis of the Tax–HTLV-1-infected population . We observed that Tax+CADM1+ cells were lysed by autologous CTL more efficiently than Tax+CADM1– and Tax–CADM1+ cells in all individuals assayed . Since the majority of the Tax+ cells are CADM1+ , this result agrees with previous observations that cells expressing the immunodominant Tax protein are efficiently lysed by autologous CTLs [12 , 43] . The Tax–CADM1+ cells were indeed lysed at a lower rate than Tax+CADM1+ cells . Because Tax controls the expression of the plus-strand encoded proteins Pol , Gag and Env , it is likely that , with the exception of HBZ , the level presentation of epitopes from other viral genes is lower in Tax−infected cells than in Tax+ infected cells . Weak immunogenicity and low levels of presentation of HBZ peptides [17] contribute to the lower rate of lysis of this population . Surprisingly , a small proportion of Tax+CADM1– cells were found to evade lysis . Thus , we propose that CADM1 facilitates CTL recognition and lysis of the infected cell and that Tax expression alone is not sufficient . Our results indicate that the efficiency of CTL killing of the Tax+CADM1+ population is predictive of the efficiency of killing of the Tax–CADM1+ population; i . e . , individuals with a higher rate of lysis of the Tax+CADM1+ cells also had a higher rate of lysis of the Tax–CADM1+ . This lower rate of clearance of the Tax-negative infected cells may be sufficient to maintain the equilibrium in vivo , because these cells do not undergo Tax-induced proliferation . As previously reported [12] , the rate of lysis parameter at a given PVL was greater in patients with HAM than in ACs . Using a CMV-specific CTL clone , we observed that CADM1+ cells , whether Tax+ or Tax– , were more susceptible than CADM1– cells to CTL-mediated lysis , even at very low concentrations of antigen . We conclude that Tax-induced host proteins make little contribution to the cells’ susceptibility to CTL-mediated lysis . The CADM1– cells , even those that express Tax , have very low susceptibility to CTL lysis unless they are loaded with artificially high concentrations of the peptide . Hence , we conclude that CADM1 is a major determinant of the susceptibility of the cell to CTL-mediated lysis . Rowan et al . ( 2014 ) reported that Tax-expressing CD4+ T cells were preferentially killed by CTLs of unrelated specificity [17] . As Tax+ cells are dominated by those of the CADM1+ phenotype , we attribute the preferential lysis of Tax+ cells in this experiment by CTLs to the presence of CADM1 on their surface . The susceptibility to lysis of the smaller Tax–CADM1+ and Tax+CADM1– populations observed in our assay are likely to have been masked by the larger proportions of Tax–CADM1– cells and Tax+CADM1+ cells respectively . CD8+ T cells , when cultured with HTLV-1-infected CD4+ T cells , expressed both CADM1 and CRTAM . Although the CD8+ T cells had a much greater frequency of expression of CADM1 than CRTAM , the frequency of CADM1 expression bore no relationship to the CTL efficiency , and in fact represents the frequency of HTLV-1 infected CD8+ T cells . However , the frequency of CRTAM expression on CD8+ T cells was associated with high rates of CTL lysis . We propose that the interaction of CADM1 with CRTAM stabilises and prolongs the contact between the infected cell and the CTL and lowers the threshold of CTL activation . To summarise , efficient CTL-mediated lysis of HTLV-1-infected cells is associated with the expression of CADM1 on the target cell and the high affinity ligand CRTAM on the effector CTL . Since CADM1 expression is tolerated in the face of enhanced CTL surveillance of HTLV-1-infected cells , we infer that CADM1 expression plays an important role in the persistence of HTLV-1 .
All donors attended the National Centre for Human Retrovirology ( Imperial College Healthcare NHS Trust , St Mary's Hospital , London ) and gave written informed consent in accordance with the Declaration of Helsinki , with the approval of UK National Research Ethics Service ( 15/SC/0089 ) . Samples were donated by 29 ACs , 1 HTLV-1+ subject with polymyositis ( P ) , 30 patients with HAM/TSP and 8 uninfected individuals ( S1 Table ) . PBMCs were isolated from whole blood by density-gradient centrifugation using Histopaque-1077 ( Sigma ) and cryopreserved in FBS ( Invitrogen ) containing 10% dimethylsulphoxide ( Sigma ) . Genomic DNA was extracted from unfixed PBMCs using the DNeasy kit ( Qiagen ) according to the manufacturer’s protocol . The lysis step was extended from 10 minutes at 56°C to 16 hours at 62°C when extracting DNA from formaldehyde-fixed cells . To quantify the PVL , a series of dilutions of genomic DNA starting from 5ng/μl was subjected to real-time quantitative PCR ( qPCR ) with the following primers: Tax: SK43 5’-CGGATACCCAGTCTACGTGT-3’ and SK44 5’-GAGCCGATAACGCGTCCATCG-3’; GAPDH: GAPDHF 5’- AACAGCGACACCCATCCTC-3’ and GAPDHR 5’- CATACCAGGAAATGAGCTTGACAA-3’ . qPCR was performed using the QuantStudio 7 Flex real-time PCR system ( Life technologies ) with the standard Fast SYBRgreen ( Life technologies ) thermal cycle protocol . A patient-derived infected CD4+ T cell clone with a mapped single integrated provirus was used as a standard reference [41] . PBMCs were analysed either uncultured or after culturing for 18 hours at a density of 1 x 106 cells/ml in RPMI-1640 ( Sigma ) containing 10% FBS , 2mM L-glutamine ( Sigma ) , 50 U/ml penicillin , 50μg/ml streptomycin ( RPMI10 , Sigma ) , supplemented with 20nM concanamycin A ( CMA , Calbiochem ) . This incubation allowed the onset of spontaneous Tax expression while CD8+ T-cell degranulation was inhibited by concanamycin A . Cells were washed once with PBS ( Sigma ) , then stained with a fixable viability dye ( Live/Dead blue or Live/Dead near infrared , Molecular Probes ) at 1μl/ml for 5 minutes . All steps were carried out at room temperature unless stated otherwise . Cells were washed with FACS buffer ( PBS 7% normal goat serum , Sigma ) . Then they were stained for 20 minutes with antibodies specific for surface markers , including CD3-Qdot605 ( clone UCHT1 , Molecular Probes ) ; CD4-eFluor450 ( RPA-T4 , eBioscience ) ; CD8-AF700 ( LT8 , AbD Serotec ) ; CCR4-PerCP-Cy5 . 5 ( TG6/CCR4 , Biolegend ) ; ICAM-1-PE ( HA58 , Biolegend ) ; CD25-APC ( M-A251 , BD ) and CADM1 ( 3E1 , MBL ) biotinylated using EZ-Link Micro Sulfo-NHS-LC-Biotinylation Kit ( Thermo Fisher Scientific ) . Unbound antibodies were removed by washing with FACS buffer . The cells were then stained for 10 min with Streptavidin-PECy7 ( 0 . 4μl/100μl , Biolegend ) . After another wash with FACS buffer , the cells were fixed with fixation/permeabilisation buffer ( FoxP3 buffer set , eBioscience ) for 30 minutes . The cells were then washed with the permeabilisation buffer and stained for the intracellular viral protein Tax for 25 minutes using the LT-4 antibody conjugated to AF488 ( Y . Tanaka ) . Finally the cells were washed and resuspended in PBS , after which they were acquired using a BD LSRFortessa . Data was analyzed using Kaluza software ( Beckman Coulter ) . The gating strategy is illustrated in S1 Fig . PBMC from twelve HTLV-1 infected donors were stained with a viability stain and antibodies specific for CD4 , CD8 , CD3 , CADM1 , CD25 and CCR4 as described . Cells were fixed with 2% paraformaldehyde for 20 min , after which live CD3+ CADM1+/–CD4+ and CADM1+/–CD8+ cells were sorted with a BD FACSAria III as outlined in S2 Fig . Genomic DNA was extracted from both unfractionated PBMC and sorted samples , and the number of proviruses present in each fraction was estimated by qPCR . The frequency CADM1+ and CADM1– cells in uncultured CD3+ cells was used to calculate the proportion of proviral load which was carried in each population of cells in each individual . CEM cells were electroporated with 2 μg of the following plasmids: GFP-Tax [48] , HBZ-IRES- GFP ( pCMV . IRES . GFP . Myc ( x2 ) _HBZ . SP1; [49] ) , and/or GFP alone ( pCMV . IRES . GFP . Myc ( x2 ) ; Clontech ) , or 1 μg HBZ-TTG ( pME18Sneo-HBZ TTG; [42] ) plus 1 μg GFP or–GFP-Tax where indicated . Electroporation was carried out using a nucleofector 1 device ( program A030 ) which routinely gave 20–40% transfection efficacy with ~95% viability . Post electroporation cells were placed in warm RPMI 10% FCS at a density of 3x105/ml in the presence or absence of 10 ng/ml phorbol 12-myristate 13-acetate ( PMA ) and 0 . 5 μg/ml calcimycin ( CAI , Sigma ) . After 16h culture , cells were harvested , stained with a viability stain and biotinylated anti-CADM1 , followed by Streptavidin-PEDazzle ( Biolegend ) . Cells were analysed by flow cytometry within 1 h . CD8+ T cells were depleted from PBMCs by magnetic cell separation ( Miltenyi Biotech ) , then mixed with the CD8-depleted fraction at a range of ratios ( including the individual’s physiological CD4+:CD8+ ratio ) and co-cultured for 18 hours at a density of 1 x 106 cells/ml in RPMI10 supplemented with 20μg/ml DNase ( Sigma ) . All samples were assayed in duplicate . Following co-culture , 10% of the cells in each tube were set aside for estimation of absolute cell counts and the remainder were stained as described above with a viability stain , anti-CD3-BV510 ( UCHT-1 , Biolegend ) , -CD4-BV605 ( RPA-T4 , Biolegend ) , -CRTAM-PE ( Cr24 . 1 , Biolegend ) , -CD8-AF700 , -CADM1–Biotin and -Tax-Cy5 . For absolute cell counts , cells were stained with CD4-BV605 and CD8-AF700 for 30 min and , without washing , fixed with 2% paraformaldehyde ( Sigma ) for 30 minutes . A fixed number of CountBright absolute counting beads ( Molecular Probes ) were added to each sample before flow cytometric analysis . Thus , the absolute number of CD4+ and CD8+ T cells per tube was calculated and used to calculate the rate of clearance of target CD4+ T cells per percentage CD8+ T cells per day , as described in Asquith et al 2005 [12] . A cytomegalovirus ( CMV ) -specific CD8+ T cell clone ( a gift from Tao Dong ) that recognises an HLA-A*0201-restricted peptide , PP65 ( NLVPMVATV ) , was expanded by co-culturing with gamma-irradiated ( 3000 rads ) PBMCs in RPMI10 containing 50 μg/ml phytohemagglutinin ( Roche ) and 100 IU/ml IL-2 ( Promocell ) . The cells were fed with fresh medium and IL-2 every 3–4 days . The CTLs were used in the susceptibility assay on day 17 post stimulation . CD4+ T cells were isolated from HLA-A*0201+ PBMCs using magnetic cell separation ( Miltenyi Biotech ) and cultured for 6 hours ( 1 x 106 cells/ml in RPMI10 containing 20μg/ml DNase ) , to allow Tax expression . The cells were then loaded with the CMV peptide pp65 ( Think Peptides ) at a range of concentrations ( 0–2 μM ) . CMV-specific CTLs were added to the peptide-loaded CD4+ cells at an effector:target ratio of 1:1 and co-cultured for 12 hours . All samples were assayed in duplicate . A small portion of cells was used to determine absolute cell counts and the remainder were stained with a viability stain , anti-CD3-BV510 , -CD4-BV605 , -CRTAM-PE , -CD8-AF700 , -CADM1–Biotin and -Tax-Cy5 .
|
Human T cell lymphotropic virus-1 ( HTLV-1 ) infects white blood cells ( CD4+ T cells ) for the lifetime of the host . The immune response limits viral spread , and people with a weak immune response have a high risk of developing an aggressive blood cancer , or a condition involving irreversible spinal cord damage . Virus and host are engaged in a constant battle: virus proteins drive the host cell to divide or infect new cells . We know that the viral protein Tax is an important target of the immune response , and cells which produce Tax are killed quickly . Infected cells which do not produce Tax are difficult to detect , so we have no idea how quickly they are killed . In this paper we show that most infected cells have a host protein ‘CADM1’ on their surface . We measured killing of CADM1 cells and saw that Tax+CADM1+ cells are the only infected cells which are strongly targeted by the immune response . We also found that infected cells which did not have CADM1 on the surface escaped killing , showing that CADM1 aids in immune control of HTLV-1 . These findings are an important step forward in our understanding of cellular turnover and immune control in chronic infection .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"blood",
"cells",
"flow",
"cytometry",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"microbiology",
"cloning",
"retroviruses",
"viruses",
"rna",
"viruses",
"cytotoxic",
"t",
"cells",
"molecular",
"biology",
"techniques",
"antibodies",
"research",
"and",
"analysis",
"methods",
"specimen",
"preparation",
"and",
"treatment",
"staining",
"spectrum",
"analysis",
"techniques",
"white",
"blood",
"cells",
"immune",
"system",
"proteins",
"animal",
"cells",
"proteins",
"medical",
"microbiology",
"htlv-1",
"t",
"cells",
"microbial",
"pathogens",
"molecular",
"biology",
"immune",
"response",
"spectrophotometry",
"cytophotometry",
"biochemistry",
"cell",
"staining",
"cell",
"biology",
"viral",
"pathogens",
"physiology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"organisms"
] |
2016
|
CADM1/TSLC1 Identifies HTLV-1-Infected Cells and Determines Their Susceptibility to CTL-Mediated Lysis
|
Human cancer genomes are highly complex , making it challenging to identify specific drivers of cancer growth , progression , and tumor maintenance . To bypass this obstacle , we have applied array comparative genomic hybridization ( array CGH ) to zebrafish embryonal rhabdomyosaroma ( ERMS ) and utilized cross-species comparison to rapidly identify genomic copy number aberrations and novel candidate oncogenes in human disease . Zebrafish ERMS contain small , focal regions of low-copy amplification . These same regions were commonly amplified in human disease . For example , 16 of 19 chromosomal gains identified in zebrafish ERMS also exhibited focal , low-copy gains in human disease . Genes found in amplified genomic regions were assessed for functional roles in promoting continued tumor growth in human and zebrafish ERMS – identifying critical genes associated with tumor maintenance . Knockdown studies identified important roles for Cyclin D2 ( CCND2 ) , Homeobox Protein C6 ( HOXC6 ) and PlexinA1 ( PLXNA1 ) in human ERMS cell proliferation . PLXNA1 knockdown also enhanced differentiation , reduced migration , and altered anchorage-independent growth . By contrast , chemical inhibition of vascular endothelial growth factor ( VEGF ) signaling reduced angiogenesis and tumor size in ERMS-bearing zebrafish . Importantly , VEGFA expression correlated with poor clinical outcome in patients with ERMS , implicating inhibitors of the VEGF pathway as a promising therapy for improving patient survival . Our results demonstrate the utility of array CGH and cross-species comparisons to identify candidate oncogenes essential for the pathogenesis of human cancer .
Rhabdomyosaroma ( RMS ) is the most common soft tissue sarcoma of childhood [1] and falls into two major histopathologic subtypes in children - embryonal and alveolar . Embryonal rhabdomyosaroma ( ERMS ) accounts for approximately 60% of childhood cases and is frequently associated with RAS pathway activation [2]–[5] . Treatment for either RMS subtype requires surgical resection , chemotherapy , and radiation with overall poor prognosis for patients with high-risk features , metastasis , or relapse disease . Thus , there is great interest in elucidating key molecular pathways and genetic factors that are involved in continued RMS growth and tumor maintenance . Cytogenetic studies , including array Comparative Genomic Hybridiation ( array CGH ) , identify frequent but inconsistent gains and losses of whole or partial chromosome arms and rare focal high-level amplifications in both human ERMS and ARMS [5]–[9] , largely precluding the identification of specific drivers of cancer in this disease . Moreover , array CGH and cross-species comparisons between mouse and human RMS have largely failed to identify functionally important genes contained within common copy number alterations ( CNAs ) . In one report , RMS that arose in Ptch1+/− Blmtm3Brd/tm3Brd ( a hypomorphic Blm allele ) mice exhibited a gain of chromosome 10 in 80% of cases [10] , but the oncogenes associated with this chromosomal gain remain undefined due to the large number of candidate genes found within this region . Moreover , extension of these findings to human RMS has not been reported . Rubin et al . recently showed that greater than 30% of ERMS arising in mice that harbor p53 homozygous deletion and/or Ptch1 heterozygous deletion lack a defined molecular signature or genetic lesion , suggesting undiscovered pathways likely contribute to ERMS transformation , growth , and tumor maintenance [11] . To date , there remains a need for novel gene discovery methods to identify genes and pathways essential for tumor growth , progression , and maintenance in human cancer – including ERMS . Zebrafish cancer shares molecular and pathological similarities to human disease [4] , [12]–[16] . For example , Lam et al . ( 2006 ) was the first to use comparative analysis of microarray data from zebrafish and human liver tumors to demonstrate a conserved molecular profile during tumor progression [13] . Building on this work , microarray gene expression studies of zebrafish ERMS and cross-species comparison to human disease identified RAS pathway activation as a common initiating event in zebrafish and human ERMS . Activating RAS mutations have also been identified in numerous studies of human ERMS [2]–[5] , [17] . Most recently , Paulson et al reported that 11 of 26 ( 42% ) human ERMS samples harbored activating RAS mutations along with acquisition of additional CNAs as detected array CGH [5] , suggesting that additional genetic lesions are likely required to drive oncogenic transformation to ERMS . Not surprisingly , zebrafish cancers also exhibit recurrent chromosomal gains and losses similar to those found in human cancer . For example , transgenic models of zebrafish melanoma , T-cell acute lymphoblastic leukemia ( T-ALL ) , and ERMS contain genomic imbalances including high-level gains and losses [18] . However , specific driver events could not be identified in these studies due to the low resolution of this platform . Using high-resolution array CGH , Zhang et al ( 2010 ) demonstrated the aneuploid nature of zebrafish malignant nerve sheath tumors ( MPNST ) , a feature that also characterizes the human disease , and identified a subset of genes that are co-amplified as high-copy gains in human MPNST [19] . High-resolution array CGH has also been applied to zebrafish T-ALL and identified a subset of genes contained within CNAs that were also amplified or deleted in human disease [20] . These latter two studies have demonstrated the utility of array CGH technology in detecting copy number aberrations and candidate driver genes in zebrafish tumor models , yet functional relevance of identified genes in human disease has not been reported nor have these genes been assessed for roles in regulating tumor maintenance – providing novel targets for therapy in established tumors . Capitalizing on a zebrafish model of kRASG12D-induced ERMS that shares common histopathological , genetic , and molecular characteristics of human ERMS [4] , [21] , [22] , we have utilized high-resolution array CGH to identify novel CNAs in ERMS . Remarkably , our array CGH analysis revealed focal CNAs that span short genomic regions and contain only 1–3 genes . To validate the functional significance of amplified genes in human ERMS , we prioritized six genes for initial characterization in human ERMS cell lines . Of these six genes , gene knockdown of Cyclin D2 ( CCND2 ) , Homeobox C6 ( HOXC6 ) , PlexinA1 ( PLXNA1 ) inhibited proliferation of human ERMS . PLXNA1 also exhibited important roles in blocking ERMS cells in early stages of muscle differentiation , enhancing migration , and altering anchorage-independent growth . CCND2 , HOXC6 , PLXNA1 , and Vascular Endothelial Growth Factor A ( VEGFA ) were also highly expressed in a large fraction of human primary RMS , supporting prominent roles for these genes in rhabdomyosarcoma . Chemical inhibition of VEGF signaling reduced tumor growth in vivo with an associated decrease in angiogenesis , implicating VEGF inhibitors as promising therapeutic agents for ERMS . Taking advantage of tractable features of zebrafish cancer genomes such as smaller CNA intervals and regions of conserved homology with human disease , our study demonstrates the effective use of array CGH to identify oncogenes required for continued tumor growth in human rhabdomyosaroma , providing novel therapeutic targets for the treatment of ERMS .
Array CGH was performed on genomic DNA isolated from twenty kRASG12D-induced zebrafish ERMS and compared to adjacent normal tissue . Array CGH revealed a complex CNA pattern with relative gains being observed more frequently than losses . For example , we identified 190 regions of amplification and 35 deletions recurrent in ≥3 zebrafish tumors analyzed ( Table S1 ) . Remarkably , only 2 of 20 zebrafish samples exhibit evidence of aneuploidy , contrasting starkly with human ERMS where nearly all human RMS harbored regions of extensive aneuploidy [5] . While 10 zebrafish ERMS showed evidence for CNAs in coding regions of the genome , only 3 exhibited a high frequency of multiple gains ( Table 1 and S1 ) . In total , we identified 19 gains and 2 losses in gene-containing amplicons that were recurrent in at least three zebrafish ERMS samples . Candidate genes in these regions were predominantly amplified as low-level gains , which averaged 1–3 genes and spanned only 48+/−27 kb ( +/− SD , Table 1; Fig . 1A ) . Copy number alterations were validated by qPCR of genomic DNA ( Fig . S1 ) . To assess whether CNAs identified in zebrafish ERMS were conserved in human disease , zebrafish array CGH data was compared to the high-resolution array CGH data from 26 human ERMS samples [5] . Of the 26 samples , 11 carried activating RAS mutations as assessed by Sanger sequencing analysis [5] . The two regions of chromosomal loss in zebrafish ERMS contained zebrafish-specific genes that failed to have human homologues ( Table 1 ) . By contrast , genes contained within the 19 CNA gains mapped to 21 distinct homologous genomic regions in human . Utilizing the same statistical algorithms and threshold settings as outlined by Paulson et al [5] , we discovered that 18 of 21 homologous regions were also amplified as low-copy gains in human ERMS samples ( Table 1 ) . Like zebrafish ERMS , these CNAs were focal , low-copy gains spanning 328 kB+/−278 kB and contained 7 . 3 genes on average ( Table 1 and S1 ) . To demonstrate the efficacy of our array CGH approach to identify evolutionarily conserved oncogenes essential for driving tumor progression and maintenance , we prioritized CNAs that contained genes and were amplified in both zebrafish and human ERMS . In total , six candidate genes were prioritized for further study in human ERMS based on the following criteria . 1 ) Candidate oncogenes were differentially expressed in human ERMS compared to ARMS and/or normal muscle as assessed by microarray gene expression studies . 2 ) Genes that have known oncogenic activity in other cancer types , but yet ascribed functional roles in ERMS , were prioritized for additional study to serve as “proof of concept” genes in our cross-species comparative study . 3 ) A subset of genes was selected which have unknown function in ERMS and represent potential novel oncogenes . 4 ) Amplified CNA regions that harbored a single human homologue were also prioritized . Based on these criteria , CCND2 , HOXC6 , PLXNA1 , VEGF , BRAF and LIMK1 were selected for further study ( Table 1 ) . CCND2 , PLXNA1 , VEGFA , and LIMK1 were the single genes contained within the amplified CNA intervals in human disease . BRAF was the only gene in the interval that was overexpressed in human ERMS , whereas CRY1 and TNNT1 identified within the same amplified interval were not differentially expressed when comparing human ERMS to normal muscle ( Fig . S2 ) . HOXC6 has been reported to be highly expressed in human ERMS compared with ARMS [23] , suggesting an possible role in modulating tumor growth . While CCND2 , a cell cycle regulator , VEGFA , an essential regulator of angiogenesis in a variety of cancer types and BRAF , an oncogene in a variety of cancers , most likely serve as our “proof of concept” genes for demonstrating functional significance in human ERMS . LIMK1 , HOXC6 and PLXNA1 represent potential novel candidate genes for driving tumor growth of ERMS . The six candidate genes were first assessed for anti-proliferative effects in human RD and SMS-CTR ERMS cell lines by siRNA knockdown , establishing a role for these genes in continued tumor growth and maintenance . Importantly , each of these human ERMS cell lines contains mutationally-activated RAS alleles , mimicking the zebrafish model . Effective gene knockdown was validated by quantitative RT-PCR and/or Western analysis ( Fig . 2 A and Fig . S3 B and S4 A ) . A quantitative VEGFA ELISA confirmed lower levels of secreted VEGFA in the growth medium from cells transfected with VEGFA siRNA ( Fig . S3 A , p<0 . 05 ) . Gene-specific siRNA knockdown of CCND2 , HOXC6 and PLXNA1 resulted in reduced cell viability as assessed by a luminescent cell viability assay in both cell lines ( Fig . 2 B and S4 B–H ) . By contrast , knockdown of BRAF , LIMK1 , or VEGFA failed to alter viability and/or growth in both cell lines ( Fig . 2 B and 2 C and data not shown ) . Following validation of growth effects using two additional siRNAs for CCND2 , HOXC6 and PLXNA1 in RD and SMS-CTR cell lines ( Fig . 2 C and S4 B–H ) , these genes were prioritized for additional functional studies . For example , knockdown of CCND2 , HOXC6 and PLXNA1 resulted in reduced EDU incorporation when compared to cells transfected with control siRNA in both RD and SMS-CTR cell lines , suggesting that inhibition of cell growth resulted from a block in proliferation ( Fig . 2 D ) . Apoptosis was not altered by gene knock down as assessed by Annexin V staining ( Fig . S5 ) . In total , our data uncovered important roles for CCND2 , HOXC6 and PLXNA1 in regulating ERMS proliferation , validating the role of several novel genes in regulating continued tumor cell proliferation in human ERMS cells . ERMS expresses myogenic factors such as MYOD and MYF5 yet it fails to complete normal myogenesis secondary to differentiation arrest [24] , [25] . As a result , ERMS is composed of heterogeneous subpopulations of proliferating tumor cells that vary in their differentiation status . Therefore , oncogenes that are essential for regulating proliferation of ERMS cells likely also play a role in modulating their differentiation status . Thus , we determined whether CCND2 , HOXC6 , and PLXNA1 also played a role in blocking differentiation of ERMS . Knockdown of PLXNA1 resulted in increased formation of multinucleated myocytes and induction of myosin heavy chain expression in RD cells ( Fig . 3 B , p = 0 . 03 ) . By contrast , siRNA knockdown of CCND2 and HOXC6 did not alter the differentiation status of human RD cell ( Fig . 3 E ) . To validate the phenotype of PLXNA1 knockdown , two independent PLXNA1 shRNA knockdown stable lines were generated and cultured under differentiation condition . Both PLXNA1 shRNAs induced robust gene knockdown compared to control scrambled shRNA ( Fig . 3 F ) , resulting in increased numbers of multinucleated-myocytes and myosin heavy chain expression ( Fig . 3 D–E , p = 0 . 01 ) . PLXNA1 also played a critical role in regulating anchorage-independent growth in colony formation assays . Stable knockdown of PLXNA1 resulted in impaired anchorage-independent growth with decreased colony formation two-fold over 15 days when compared to RD cells transduced with control shRNA ( Fig . 3 G–I , p = 0 . 0003 ) . Moreover , colonies were smaller in size , likely reflecting the prominent role of PLXNA1 in regulating cell growth . Together , our findings indicate that PLXNA1 plays an essential important role in regulating proliferation and differentiation in transformed ERMS . Migratory behavior of tumor cells in vitro can be a useful predictive index of cell invasion and metastasis in vivo . Genes and pathways that are essential for regulating the migratory behavior of tumors cells can likely serve important functions in mediating metastasis and therefore are potential targets for novel therapy . Wound healing and transwell migration assays were used to assess a role for CCND2 , HOXC6 , PLXNA1 and VEGFA in migration of human RD and SMS-CTR ERMS cell lines . Knockdown of PLXNA1 by siRNAs ( smart-pool and individual siRNAs ) and shRNAs resulted in impaired migration in both RD and SMS-CTR cells over 22 hours ( p<0 . 02 for RD and p≤0 . 04 SMS-CTR , Fig . 4 G–I , Fig . S6 ) . By contrast , knockdown of CCND2 , HOXC6 and VEGFA did not affect migration of either RD or SMS-CTR cells ( p>0 . 25 , Fig . 4 A–F , I ) . As an independent assessment of ERMS cell migration , PLXNA1 stable shRNA knockdown cells were assessed for migration in a transwell assay . Knockdown of PLXNA1 in RD cells with two independent gene-specific shRNAs resulted in >50% reduction in transwell migration ( p = 0 . 03 for shRNA-A and p = 0 . 0038 for shRNA-B , Fig . 4 J ) . Together , these results support an additional role for PLXNA1 in regulating migratory behavior of human ERMS cells . The VEGFA pathway often exerts powerful roles in regulating cancer-induced angiogenesis , which would have been missed in our human cell culture assays . To assess a role for VEGFA in modulating tumor growth in vivo , ERMS-bearing zebrafish were treated with the VEGF receptor tyrosine kinase inhibitor , cediranib , or DMSO vehicle for 7 days and assessed for effects on tumor growth . Relative tumor growth as determined by the ratio of tumor volume change between pre- and post-treatment was reduced by three-fold in cediranib-treated fish when compared to those treated with vehicle ( Fig . 5A–M , p = 0 . 0017 , Student's T-test ) . As VEGFA is known to promote angiogenesis during tumor progression in a variety of cancers , we next assessed whether inhibition of VEGFA also blocked angiogenesis in ERMS in vivo . In order to visualize angiogenesis in established tumors , ERMS co-expressing rag2-KRASG12D and rag2-dsRED were transplanted into irradiated fli1-GFP fish that exhibit vessel-specific GFP expression [26] . Fish with engrafted ERMS were treated with either cediranib or DMSO vehicle for 7 days . Animals were assessed for differences in both overall tumor growth and microvessel density as determined by cryosections of tumors . ERMS-affected animals treated with cediranib showed a significant reduction in tumor growth with an accompanied two-fold reduction in tumor microvessel density when compared to those treated with vehicle control ( N = 5 for each group , p = 0 . 006 , Fig . 5 N–P ) . Cediranib-treated ERMS did not exhibit a difference in proliferation when compared to vehicle control-treated tumors ( Fig . 5 Q–S ) , consistent with our results for VEGFA gene knockdown in human ERMS cell lines . Together , these data suggest that activation of the VEGF pathway promotes ERMS tumor progression through enhanced angiogenesis . Having established roles for CCND2 , HOXC6 , PLXNA1 and VEGFA in ERMS growth , we next wanted to assess the extent to which these proteins are expressed in human primary RMS . Immunohistochemistry was performed using antibodies to CCND2 , HOXC6 , PLXNA1 and VEGFA in primary human tumors and fetal muscle ( Supplemental Table S2 ) . In all , 8 pediatric and 11 adult ERMS and 3 pediatric and 4 adult alveolar RMS ( ARMS ) were analyzed . Remarkably , CCND2 , HOXC6 , PLXNA1 and VEGFA protein expression were detected in a majority of RMS samples while antibody staining for each was largely negative in fetal muscle ( Fig . 6 ) . Specifically , HOXC6 protein expression was detected in 14 of 19 ERMS with strong , diffuse staining being found in 6 of the 14 cases ( 1 adult and 5 pediatric ) . In contrast , only 2 of 7 ARMS showed weak , positive staining for HOXC6 , consistent with lower-level gene transcript levels being detected in pediatric ARMS compared to ERMS ( Supplemental Fig . S7 ) . CCND2 , PLXNA and VEGFA were expressed at comparable frequency in both subtypes of RMS . For example , CCND2 was detected in 15 of 19 ERMS and 5 of 7 ARMS , while PLXNA1 expression was found in 17 of 19 ERMS and 6 of 7 ARMS . VEGFA antibody staining was detected in the tumor cells and the vasculature in 10 of 19 ERMS ( strong staining in 1 adult and 1 pediatric case ) , while 6 of 7 ARMS exhibited weak staining in all cases analyzed . Additional immunohistochemical analysis of a tissue microarray from Children's Oncology Group revealed positive VEGF expression in 31 of 38 ERMS and 3 of 6 ARMS ( Table S3 ) . Of the 38 cases of ERMS , 29 cases showed strong and diffuse staining . Our analysis suggests that despite these oncogenes being infrequently amplified in human disease , their protein expression levels are elevated in a majority of human ERMS . These data imply important roles for these genes in regulating tumor growth in a large fraction of human ERMS and suggesting additional , as of yet undiscovered mechanisms that regulate expression of these genes . To assess whether dysregulated expression of CCND2 , HOXC6 , PLXNA1 and VEGFA correlates with clinical outcome , Kaplan Meier analyses were completed using microarray gene expression data from primary ERMS and ARMS [23] . Samples were stratified based on high and low median expression for each gene and each assessed as an independent predictor of survival . Based on this analysis , differential expression of CCND2 and PLXNA1 did not correlate with overall survival outcome in either ERMS or ARMS ( Fig . S8 ) . HOXC6 was differentially upregulated in ERMS compared to ARMS ( Fig . S7 ) ; thus , high expression of HOXC6 correlated with better overall survival ( Fig . 7 A ) , a finding consistent with previous studies demonstrating better clinical outcome for ERMS patients compared to those with ARMS [27] . Finally , samples with high mRNA expression of VEGFA correlated with low overall clinical survival in the ERMS cohort but did not predict survival outcome in ARMS ( Fig . 7 B ) . In addition , VEGFA expression did not correlate with clinical stage , indicating that it is likely an independent prognostic indicator ( Fig . S9 ) . These data implicate important roles of VEGFA in promoting ERMS tumor progression and identify VEGFA as a biomarker with likely use in stratifying ERMS patients into high and low-risk groups .
Prior cytogenetic and array CGH studies in human ERMS demonstrate inconsistent and non-specific partial to whole chromosomal aneuploidy across different primary tumors making it difficult to identify critical genes essential for driving tumor growth . Utilizing a zebrafish model of RAS-induced ERMS that mimics the human disease [4] , [21] and subsequent array CGH analyses of genomic DNA from tumor vs . matched normal , we were able to rapidly identify candidate gene-containing regions that likely contribute to ERMS pathogenesis . The 19 CNA gains that were recurrently amplified in zebrafish ERMS mapped to 21 homologous regions within the human genome . Remarkably , 18 of these regions also demonstrated low-level genomic amplification in human ERMS . To validate that candidate genes contained within these intervals exert important roles in continued tumor growth and maintenance , we characterized the function of six amplified genes in human ERMS cell lines and conclusively demonstrated functional significance of CCND2 , HOXC6 , PLXNA1 in proliferation of human ERMS . PLXNA1 also has important roles in regulation differentiation and migration of ERMS cells . As the in vitro analyses performed in this study would not be able to assess other aspects of tumorigenesis such as neovascularization and tumor initiation , we utilized the zebrafish in vivo model to demonstrate the important role of VEGF-A pathway in mediating angiogenesis during tumor growth . In total , our work identified roles for 4 of 6 candidate genes identified in our cross-species array CGH studies for eliciting important roles in human ERMS . Importantly , this strategy is not limited to zebrafish ERMS , and will likely provide powerful new methods to identify novel tumor-suppressor and oncogenes in a wide range of zebrafish and human tumors . Data from our array CGH study and previous studies of zebrafish cancer revealed low-level CNA gains as a frequent DNA alteration in cancer , yet this class of mutation has not commonly been studied due to the difficulty in identifying relevant and meaningful genes in these regions . Importantly , zebrafish allows for the easy identification of low-level gene amplifications . In total , our data is consistent with a model where zebrafish tumor cells undergo acquisition of low-amplitude gains , likely represented as single copy gains within CNA regions . For example , we have also observed that clonal-populations of purified T-ALL cells ( 90% enriched for blasts ) also contain low-amplitude gains [Blackburn et al . , unpublished] . Moreover , Rudner et al . ( 2011 ) recently showed that a majority of amplified , gene-containing CNAs found in zebrafish T-ALL were also amplified in human disease [20] . Upon re-analysis of this data , we find that 72% of the reported amplified regions were detected as low-level gains in zebrafish T-ALL , yet were not reported as such . Zhang et al . identified large regions of aneuploidy and high-level CNA gains in zebrafish malignant peripheral nerve sheath tumors when assessed by array CGH , but also identified numerous regions of low-level CNA gains , which were dismissed as potential causative lesions in cancer . Thus , despite these previous two reports observing low-level CNA gains in zebrafish malignancy , neither reported the functional importance of this class of genes to promote tumor progression and maintenance in zebrafish or human disease . Although it is formally possible that low level gains detected in zebrafish ERMS represent high-copy gains masked by a high degree of tumor cell heterogeneity and/or contamination of normal DNA from non-transformed blood , fibroblasts and stroma , our data strongly argue that low-copy amplification is a common attribute found in zebrafish and human cancer . Interestingly , even though the functional relevance of low-level gains such as genomic duplication events have been infrequently reported in human cancer , this type of DNA alteration often predicts important clinical parameters such as disease susceptibility , therapy resistance and adverse prognosis . For example , duplication of a region on chromosome 6q27 is detected in individuals affected with familial chordoma , a rare bone cancer , but not among unaffected individuals within the same family [28] . MYB tandem duplication occurs in pediatric T-ALL and results from homologous recombination at ALU repetitive sequences flanking the MYB locus . Elevated MYB expression is associated with poor outcome in T-ALL [29] . Similarly , focal tandem duplication also contributes to chemotherapy resistance in patients with high-grade ovarian cancer [30] . These findings indicate that low-level CNA gains have important clinical prognostic relevance and likely play important functional roles in human cancer . Finally , we also found that genes within each CNA are highly expressed in a majority of human RMS despite being infrequently amplified as low-copy CNAs , suggesting the importance of these gene pathways in regulating a large fraction of human ERMS and that additional mechanisms underlying the dysregulation of this class of genes in cancer is likely . Our work has identified essential roles for four genes in modulating ERMS growth , maintenance , migration , and neovascularization . Of these genes , CCND2 , HOXC6 and PLXNA1 exhibited important roles in regulating proliferation in human ERMS cell lines . PLXNA1 also had additional roles in arresting ERMS cells in early stages of muscle differentiation , in enhancing tumor cell migration , and in altering anchorage-independent growth . Despite the fact that these genes and/or related family members have been ascribed functions in other cancer types , their contributions to the pathogenesis of ERMS have not been previously characterized . For example , HOXC6 , a homeobox transcription factor , regulates the expression of genes including BMP7 , FGFR2 , IGFP3 and PDGFRA to influence oncogenic activities in prostate cancer [31] . HOXC6 is highly expressed in ERMS but not ARMS [23] , suggesting a specific and independent role in regulating growth in the human ERMS subtype . A role for HOXC6 in regulating continued RMS growth had not been reported until this study . CCND2 belongs to the D-type G1 cyclins ( D1 , D2 and D3 ) . While cyclin D1 is frequently dysregulated in cancer and is a marker for disease progression [32] , the involvement of cyclin D2 in cancer is not as well characterized . CCND2 is amplified in 2% of gliomas and in zebrafish and human MPNSTs [19] , [33] . Finally , PLXNA1 belongs to a highly conserved family of transmembrane receptors that bind semaphorins and have been shown to mediate neuronal cell migration , guidance , and patterning [34] , [35] . In humans , nine plexins group into four subfamilies and several have been implicated as having roles in cancer progression and growth . In particular , plexin-B1 can function as an oncogene by promoting proliferation and survival of B-Cell Lymphoblastic Lymphoma cells and invasion of ovarian and breast tumor cells [36]–[38] . Plexin-A1 , the gene identified in our study , has been shown to activate the VEGF receptor and NF-κB to promote survival of malignant mesothelioma cells [39] , suggesting a complex interplay of PLXNA1 in cell survival and neovascularization . Taken together , our study has demonstrated prominent and novel roles for CCDN2 , HOXC6 , and PLXNA1 in modulating ERMS proliferation while PLXNA1 exerts important additional roles in regulating differentiation and migration . None of these genes have been previously implicated as important modulators of ERMS growth and maintenance , suggesting that our cross species array CGH studies will be valuable for uncovering genetic lesions across a wide range of zebrafish and human cancers . VEGF pathway activation promotes tumor angiogenesis and progression in a variety of human cancers , and elevated VEGF expression correlates with poor prognosis in certain tumor types [40]–[42] . However , until our report , the prognostic impact of VEGFA expression in human ERMS had not been described . Here , we show that VEGFA is amplified as a low-copy gain in a small cohort of zebrafish and human RMS and yet highly expressed in a majority of human patient samples . High VEGFA mRNA expression correlated with poor clinical outcome in human ERMS , underscoring the importance of this pathway in driving continued tumor growth . As VEGFA expression level is not linked to clinical stage , it represents an important independent prognostic indicator and a potential biomarker for therapy stratification . Chemical inhibition of VEGF signaling in our pre-clinical in vivo model effectively suppressed tumor growth by reducing angiogenesis , consistent with the findings from a pre-clinical testing of VEGFR inhibitors on a small number of human RMS xenografts into mice [43] . Although clinical trials of VEGF inhibitors in other types of cancers have exhibited mixed results [44]–[46] , our data suggest that targeting the VEGF pathway may be a promising therapeutic option to curb tumor growth in a subset of high-risk ERMS patients . In summary , our array CGH studies of zebrafish cancer have identified conserved CNA gains with functional significance in human ERMS . As proof of principle , we have also demonstrated the utility of zebrafish array CGH studies to identify oncogenes that are essential for continued tumor growth in both zebrafish and human ERMS . Our work also provides 13 additional CNA gains that are conserved in zebrafish and human ERMS for which an essential genetic lesion has yet to be identified – providing potential genes to interrogate in the future . Our studies suggest that most amplified CNAs will contain genes that regulate important processes in cancer maintenance and growth . Moreover , our study reveals a number of tractable features of zebrafish cancer genomes such as small-size CNAs containing few genes within each region of chromosomal aberration , thereby positioning the zebrafish as an effective model system for discovering novel genes required for continued tumor growth and maintenance within a wide diversity of cancer types .
Studies were approved by the Massachusetts General Hospital Subcommittee on Research Animal Care under protocol #2011N000127 ( zebrafish ) and by the Partners Human Research Committee under IRB protocol #2009-P-002756 ( human ) . TuAB-strain zebrafish were co-injected at the one-cell stage with linearized rag2-KRASG12D and rag2-dsRED DNA constructs as previously described [4] , [47] . dsRED-labeled ERMS and adjacent non-neoplastic tissues were dissected from tumor-bearing animals at 30–40 days of life . RNA and DNA were extracted by Trizol ( Sigma ) . Tumor DNA was labeled with Cy5 ( Bioprime system , Invitrogen , Carlsbad , CA ) and hybridized against the matched normal samples labeled with Cy3 onto the custom SurePrint G3 400k CGH microarray ( Agilent Technologies , Santa Clara , CA ) . Array image scans were extracted using Agilent Feature Extraction software ( Agilent Technologies , Inc , Santa Clara , CA ) , normalized for signal intensity , and imported into the Nexus Copy Number software program version 5 . 1 ( Biodiscovery , Inc . , El Segundo , CA ) . CNA calls were generated based on log2 ratio output files using a rank segmentation algorithm . Settings were optimized using self-self hybridizations to reduce false positive calls . The parameters include significance threshold 1 . 0 E-5 , maximum continuous probe spacing of 200 kb , minimum number of probes per sequence of 3 , and log2 ratios of 1 . 0 , 0 . 25 , −0 . 25 and −1 . 0 for high-level amplifications , gains , losses and deletions , respectively . CNAs of interest were determined using the aggregate function in Nexus . Aggregates are represented as segmented regions of gain or loss shared by a set of samples with the number of samples sharing the event referred to as the aberration frequency . The minimum aberration frequency required for analysis in our study was set at ≥15% ( n≥3 of 20 zebrafish ERMS contained a common region of gene amplification ) . For the human ERMS sample analysis , normalized log2 intensity files ( series number GSE27392 ) were downloaded from Gene Expression Omnibus ( GEO ) at the National Center for Biotechnology Information ( NCBI ) and imported into and analyzed using Nexus Copy Number software ( version 5 . 1 , BioDiscovery ) . This program analyzes log2 ratio output files using a rank segmentation algorithm similar to circular binary segmentation . Samples were segmented following the removal of the greatest 3% of outliers and a minimum five-probe requirement at a significance threshold of 1E-08 . Gains and losses were defined as regions exhibiting log2 values of 0 . 2 and −0 . 18 , respectively , with high-level amplifications and deletions defined as log2 values greater than 0 . 5 and less than −0 . 5 . Following the identification of human ERMS aberrations , homologous human regions of zebrafish ERMS CNAs were analyzed to determine whether common low-level amplifications were present in both zebrafish and human ERMS samples . Paraffinized human primary rhabdomyosaroma ( 5 ERMS and 3 ARMS ) , US Biomax tissue microarray ( 14 ERMS and 4 ARMS , Rockville , MD ) , and a Children's Oncology Group tissue microarray were analyzed by immunohistochemical staining as previously described [48] . HOXC6 ( Sigma , 1∶200 ) , CCND2 ( Ab-Cam , 1∶100 ) , PLXNA1 ( Ab-Cam; 1∶200 ) and VEGFA ( Ab-Cam; 1∶125 ) . BGAR- biotinylated goat anti-rabbit ( Vector #BA-1000 ) was used as the secondary antibody . Pathology review was completed independently by E . C . and G . P . N . The human RD cell line was obtained from ATCC cell biology collection ( Manassas , Virginia ) and the SMS-CTR cell line provided by Dr . Corrine Linardic ( Duke University , North Carolina ) . Cells were seeded at a density of 5×102 cells in 6-well plates in 2 ml of antibiotic-free 10% FBS/DMEM . 50 pg of gene-specific smart-pool or control siRNA were transfected into cells using RNAiMax lipofectamine transfection reagent ( Invitrogen ) . For stable knockdowns , scrambled and gene-specific shRNAs in pLKO . 1-based lentiviral vectors were packaged in 293T cells . shRNAs were obtained from molecular profiling laboratory at the Cancer Center of Massachusetts General Hospital ( Table S4 ) . RD cells were infected with viral particles for 24 hours at 37 degrees with polybrene ( Millipore ) at 4 µg/mL and then selected with puromycin ( In Vivo Gene ) at 10 µg/mL in 10%FBS/DMEM for 15 days to obtain stable lines . Total cell lysates from knockdown experiments were immunoblotted using primary antibodies against HOXC6 ( 1∶500 ) , CCND2 ( 1∶1000 ) , PLXNA1 ( 1∶1000 ) and VEGFA ( 1∶1000 ) . All Western analysis was completed three times per experiment and average percent knockdown is noted . Incubation with HRP-conjugated secondary antibody ( 1∶2000 ) was performed in 5% milk/TBST for 2 hours . siRNA transfected cells were assessed by Cell Titer Glo assay as per the manufacturer's instructions ( Promega ) . Cells were also pulsed with EDU for 2 hours , harvested at 48 hours post-transfection and processed using the EDU ClickIt Flow Cytometry Assay kit ( Alexa Fluor 647 dye , Invitrogen ) . Unstained cells were used as the negative sample to facilitate gating in flow cytometry . To assess apoptosis , cells were harvested at 48 hours post-transfection and labeled with PE Annexin V and 7-AAD using the PE Annexin V Apoptosis Detection Kit ( BD Pharmagin ) . Unstained cells , cells treated with PE Annexin V only and 7-AAD only were used to set up gates for flow cytometry . Each analysis was performed in triplicate . A student's T-test was performed to assess whether the difference in the percentage of Annexin V-positive cells between test samples and control siRNA-transfected cells was significant . A wound-healing assay was performed in cells transiently transfected with siRNA and/or cells that stably express a gene-specific shRNA . Cell were seeded into 6-well plates and grown to nearly confluent density . A scrape was made in each well using a pipette tip , and cell migration across the gap was assessed after 22 hours . Images were taken at 0 and 22 hrs to calculate the percentage of gap closure . ERMS cells were also analyzed for altered migration in a transwell assay . Specifically , 2×104 cells were seeded in 6 . 5 mm-membrane inserts ( Corning ) in DMEM and were allowed to migrate through the permeable membranes ( 8 . 0 µM pore size ) toward the bottom chamber containing medium with 10% FBS . Cells were then fixed with 4% paraformaldehyde after 24 hours and stained with hematoxylin for 30 minutes . Unmigrated cells from the inserts were removed . Six random fields of the migrated cells on the membranes were imaged using the Olympus light microscope ( Model MVX10 , 400× magnification ) and manually counted . A Student's T-test was performed to assess differences between the control and experimental groups . A base layer of 1% agar in 10% FBS/DMEM was prepared in 6-well plates . Cells were resuspended in 0 . 5% low-melting point agarose/10%FBS/DMEM and overlaid on the base layer with 2 . 5×103 cells per well and subsequently kept in the humidified incubator with 5% CO2 with media change every 3 days for 15 days . Cells were fixed with 4% paraformaldehyde and stained with 0 . 05% crystal violet . Colony formation assay for each shRNA stable knockdown line was performed in triplicate . Image for each well containing soft agar colonies was taken at low magnification by light microscopy . Colony count was performed using the ImageJ software and differences assessed by Student T-test . Six-week old CG1 syngeneic fish were transplanted with 3×104 unsorted tumor cells arising from dsRED-positive ERMS from CG1 strain fish ( Mizgireuv and Revskoy 2006; Smith et al . , 2010 ) . Engrafted animals were treated at 6-days post-transplantation with 100 nM of cediranib ( Selleck ) and vehicle control ( DMSO ) for 7 days ( including 2 24-hr drug holidays ) . Tumor volume was assessed by imaging animals pre-treatment and post-treatment . Tumor volume was calculated by multiplying tumor area by fluorescent intensity using image J . A Student's T-test was performed to assess differences between tumor size in the control and experimental groups . Six-week old fli1-GFP fish were irradiated at 25 Gy and transplanted with 3×104 unsorted ERMS cells from fish with dsRED-positive ERMS . Fish with engrafted tumors were treated with cediranib as described above . At the end of treatment period , tumor tissues were isolated , fixed in 4% paraformaldehyde for 30 minutes and snap frozen . 5 µM Frozen sections were mounted in DAPI-containing Vectashield ( Invitrogen ) . GFP and dsRED images were obtained at 200× magnification using a Nikon confocal microscope . Microvessel density was quantified using Weidner et al . criteria [49] and differences assessed by Student T-test . Kaplan-Meier analysis was completed using R with the survival package . Median expression level for each gene was used to group samples into high and low expression . Chi-squared tests were used to assess overall survival differences between groups . Statistical significance was defined as a p-value less than 0 . 05 .
|
Cancer is a complex genetic disease that is often associated with regional gains and losses of genomic DNA segments . These changes result in aberrant gene expression and drive continued tumor growth . Because amplified and deleted DNA segments tend to span large regions of chromosomes , it has been challenging to identify the genes that are required for continued tumor growth and progression . Array comparative genomic hybridization ( array CGH ) is an effective technology in identifying abnormal copy number variations in cancer genomes . In this study , array CGH was used in a zebrafish model of embryonal rhabdomyosarcoma - a pediatric muscle tumor . Our work shows that the zebrafish cancer genome contains a small number of recurrent DNA copy number changes , which are also commonly amplified in the human disease . Moreover , these chromosomal regions are small , facilitating rapid identification of candidate oncogenes . A subset of genes identified in zebrafish array CGH was prioritized for functional characterization in human ERMS , identifying evolutionarily conserved pathways that regulate proliferation , migration , differentiation , and neovascularization . Our results demonstrate the broad utility of cross-species array CGH comparisons of human and zebrafish cancer and provide a much needed discovery platform for identifying critical cancer-causing genes in a wide range of malignancies .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2013
|
Cross-Species Array Comparative Genomic Hybridization Identifies Novel Oncogenic Events in Zebrafish and Human Embryonal Rhabdomyosarcoma
|
The Grb2-associated binding protein 1 ( GAB1 ) integrates signals from different signaling pathways and is over-expressed in many cancers , therefore representing a new therapeutic target . In the present study , we aim to target the pleckstrin homology ( PH ) domain of GAB1 for cancer treatment . Using homology models we derived , high-throughput virtual screening of five million compounds resulted in five hits which exhibited strong binding affinities to GAB1 PH domain . Our prediction of ligand binding affinities is also in agreement with the experimental KD values . Furthermore , molecular dynamics studies showed that GAB1 PH domain underwent large conformational changes upon ligand binding . Moreover , these hits inhibited the phosphorylation of GAB1 and demonstrated potent , tumor-specific cytotoxicity against MDA-MB-231 and T47D breast cancer cell lines . This effort represents the discovery of first-in-class GAB1 PH domain inhibitors with potential for targeted breast cancer therapy and provides novel insights into structure-based approaches to targeting this protein .
Overexpression of Grb2-associated binding protein 1 ( GAB1 ) has been observed in several human cancers , such as breast and lung cancers [1]–[4] . This protein is a substrate of several growth factors and interleukin receptors , and it is involved in the integration of different signal transductions [1]–[4] . Particularly , GAB1 mediates the activation of mitogen-activated protein kinase ( MAPK ) and phosphoinositide 3-kinase ( PI-3K ) cascades [5] , [6] . It belongs to a family of scaffolding proteins closely related to the insulin receptor substrates ( e . g . , IRS1 ) [2] . It contains an N-terminal pleckstrin homology ( PH ) domain binding to phosphatidylinositol- ( 3 , 4 , 5 ) -triphosphate ( PtdIns ( 3 , 4 , 5 ) P3 ) , tyrosine phosphorylation sites for the Src homology 2 ( SH2 ) binding and a proline-rich domain interacting with Src homology 3 ( SH3 ) domain [6] . PH domains can be subdivided into four groups based on their selective binding to phosphoinositides [7] , and GAB1 PH domain belongs to Group 1 which exhibits the strongest binding to PtdIns ( 3 , 4 , 5 ) P3 , but weak affinity and specificity to PtdIns ( 3 , 4 ) P2 or PtdIns ( 4 , 5 ) P2 [4] , [8] . Additionally , the phosphorylation of GAB1 on Y627 depends on the intracellular translocation from cytosol to membrane by binding to PtdIns ( 3 , 4 , 5 ) P3 via its PH domain [9] . Therefore , inhibition of GAB1 PH domain functions may prevent the recruitment of GAB1 to the membrane and suppress cancer cell ( e . g . , breast cancer ) proliferation and metastasis [10] . Herein , we attempt to identify novel small molecule inhibitors selectively targeting the PH domain of GAB1 and suggest that these small molecules exhibit high therapeutic potency for cancer treatment . Unfortunately , no three-dimensional ( 3D ) structure is available to date for GAB1 PH domain or any PH domain in complex with drug-like small molecules . Challenges remain for accurate structural prediction due to its low sequence identity ( <30% ) to other PH domains with known structures [11] . However , the core β-sandwich fold among PH domains is conserved [11] , making it possible to construct a reliable homology model structure of GAB1 PH domain . Here , based on the position-site specific matrixes ( PSSM ) obtained from all non-redundant PH domain structures , we performed fold recognition and homology modeling , followed by intensive structural refinement . The resulted model was then applied to high-throughput virtual screening of our unique collection of over five million drug and lead-like compounds with our in-house drug discovery workflow ( Fig . 1 ) [12] . Upon biological evaluation , five out of the initially tested 20 hits exhibited positive activities to form direct binding to GAB1 PH domain , inhibit GAB1 Y627 phosphorylation and suppress breast cancer cell proliferation with low micromolar IC50 . As is known , triple negative breast cancers are more aggressive with poor prognosis and difficult to treat clinically [13] , but our inhibitors showed high potency against these malicious cells . Therefore , this study validates the effectiveness of our in silico platform for drug discovery , and demonstrates that targeting the PH domain of GAB1 provides a promising and novel therapeutic strategy for cancer treatment .
PH domains are unique due to their conserved secondary structures and 3D folds , all with seven β-sheets and a C-terminal helix . However , the pairwise sequence identities among different PH domains are usually below 30% , and the loop regions are hypervariable in length and amino acid sequence [11] . Herein , we collected all available 34 non-redundant crystal structures of PH domains from Protein Data Bank ( PDB ) [14] and performed secondary structure-based sequence alignment using STRAP [15] . From the sequence alignment , we generated PSSMs for β1 , β2 , β3 , β6 , β7 and α1 ( presented as sequence logos in S1 Fig . ) to guide secondary structure prediction of new PH domain ( e . g . , GAB1 ) . As no reliable PSSMs for β4 and β5 were generated due to low sequence similarity , we used PSIPRED server [16] to predict these two β-sheets . S1 Fig . shows the sequence logos derived from the collected 34 PH domains , in which the size of residue indicates the relative frequency of that residue at the corresponding position . As expected , we found that most conserved residues are in the hydrophobic cores of PH domains . The residues responsible for phosphoinositide binding are generally located at β1[7] , β2[2] , β2[5] , β3[4] , β3[+1] and β7[1] ( the number in the brackets indicates the residue position at the secondary structure element ) . Predominantly , they are basic residues such as lysine and arginine . We combined these observations with PSSM and PSIPRED to predict the secondary structure of GAB1 PH domain , and found the predicted structure preserves a typical β-sandwich fold where C8-K14 , W26-L33 , V44-Y48 , R58-D61 , Q66-G71 , I84-N88 , and R92-V97 form the respective seven β-sheets , while E101-I114 forms the C-terminal α-helix ( Fig . 2 ) . However , the GAB1 PH domain is unique with: 1 ) a long β1 , 2 loop landmarked by the conserved K14 and W26 , similar to myosin X ( PDB ID: 3TFM [17] ) ; 2 ) a long β2 , 3 loop , similar to IRS1 ( PDB ID: 1QQG [18] ) ; 3 ) a long β5 , 6 loop , similar to TAPP1 ( PDB ID: 1EAZ [19] ) ; 4 ) the highest sequence identity of active-site residues ( except for β1 , 2 loop region ) to DAPP1 ( PDB ID: 1FAO [20] ) ( shadowed residues in Fig . 2 ) . Therefore , we have chosen the above four proteins as the templates for the following-up homology modeling studies . We constructed 1 , 000 homology models of GAB1 PH domain in complex with inositol-tetrakisphosphate ( IP4 ) based on the X-ray crystal structures of four aforementioned templates . After loop refinement and molecular dynamics ( MD ) simulation , we selected one reliable model in which IP4 binds stably to GAB1 PH domain with a minor fluctuation of phosphates ( RMSF<1 . 1 Å ) , shown in S2 Fig . The simulation of this model reached the equilibrium after 5 ns , as judged by the RMSD of all of the backbone atoms ( C , CA and N ) ( S2 Fig . ) . Large fluctuations of the Cα atoms were only observed in the β1 , 2 , β2 , 3 and β5 , 6 loops ( S2 Fig . ) . The quality of the lowest-energy model was assessed by QMEAN [21] , ProSA [22] and PROCHECK [23] . The Ramachandran plot showed reasonable backbone dihedral angles: 92 . 2% of the residues were in the most favored regions , and eight residues in the additional or generously allowed regions . Both the ProSA Z-score ( −4 . 04 ) and QMEAN Z-score ( −0 . 13 ) of final model were within the range as typically seen for the native proteins of the similar size ( S3 Fig . ) . In addition , the DOPE per-residue profile demonstrated a significant decrease in the DOPE scores at the β2 , 3 loop , β4 , 5 loop , β5 , β5 , 6 loop and β6 for the refined structure compared with the initial homology model ( S4 Fig . , and homology model coordinate file is available at http://imdlab . org/supporting/PLOSCompBio ) . As illustrated by Fig . 3A , the 3D model of GAB1 PH domain maintained the conserved β-sandwich folding . Similar to other Group 1 PH domains ( e . g . , Grp1 [20] and Btk [24] ) , the phosphoinositide-binding site of GAB1 was surrounded by the β1 , 2 , β3 , 4 and β6 , 7 loops . The 2-hydroxyl group of IP4 oriented towards the β1 , 2 loop , and the 3 , 4 , 5-phosphates intensively interacted with the aforementioned basic residues in the β1 , β2 , β4 and β7 . Particularly , K19 and R23 in the β1 , 2 loop formed hydrogen bonds with 5-P and 1-P , respectively ( Fig . 3B ) . This explains why GAB1 PH domain specifically binds to PtdIns ( 3 , 4 , 5 ) P3 but not PtdIns ( 3 , 4 ) P2 or PtdIns ( 4 , 5 ) P2 . Strikingly , the sequence motif NKKEFE in the β5 , 6 loop folded into an additional α-helix , as we termed α′ . This additional α-helix also occurs in phospholipase Cδ PH domain ( PDB ID: 1MAI [25] ) , and it interacts with W26 , F79 and Y95 in the β1 , 2 loop via hydrogen bonding networks and hydrophobic interactions ( Fig . 3C ) . This α′-helix was likely to stabilize the IP4-bound conformation of β1 , 2 loop , as W26A or W26C mutation impairs the PtdIns ( 3 , 4 , 5 ) P3 binding [8] . Furthermore , the motif SPP in the β1 , 2 loop formed intensive vdW interactions with the β7 and inositol scaffold ( Fig . 3C ) . Finally , GAB1 PH domain had an extra hydrophobic region ( later defined as Region II ) due to the smaller side chains of those hydrophobic residues around β6 , 7 loop compared to IRS1 . All these specific structural features intrinsically offered possibility of designing selective inhibitors against GAB1 over other PH domains , as further discussed in the ligand-induced conformational changes section . To identify novel inhibitors of GAB1 PH domain , we performed structure-based virtual screening using our MD-refined structural model . Additionally , a protein-based pharmacophore filter was derived using GRID method to select those high-throughput virtual screening hits of which the docked poses matched the pharmacophores [26] , [27] . Residues K14 , R23 , K27 , R29 , R58 and R92 were identified as the residues that favorably interact with hydrogen bond acceptors , whereas Y47 , F94 and I60 were specified as the preferential areas for hydrophobic moieties ( S5 Fig . ) . The residues responsible for PtdIns ( 3 , 4 , 5 ) P3 binding were predicted to be K14 , K27 , R29 , Y47 , K49 , R58 and R92 , consistent to the mutagenesis studies [6] , [8] . These critical residues were employed to define the protein pharmacophores to select docking poses of those 10 , 000 top ranked hits ( only based on docking scores ranging 43 . 47–101 . 39 ) from the virtual screening of over five million compounds of our in-house collection . The resulted 2 , 783 hits were subjected to cluster analysis based on their chemical diversity ( Tanimoto coefficient <0 . 65 ) , and we obtained 268 clusters and selected the best-scored hits from each cluster ( Fig . 1 ) . Upon visualizing their molecular interactions with the GAB1 PH domain , we chose 20 hits as listed in Table 1 and S1 Table . To validate our in silico identified hits , we performed three types of experimental assays to evaluate their bioactivities: direct binding to GAB1 PH domain , inhibition of Y627 phosphorylation of GAB1 , and cytotoxicity IC50 in triple negative MDA-MB-231 and T47D human breast cancer cells . Our experiments revealed that 10 out of 20 hits demonstrated submicromolar to micromolar binding affinity ( <50 µM ) to GAB1 PH domain measured by surface plasmon resonance ( SPR ) . Among them , GAB-001 , GAB-004 , GAB-007 , GAB-016 and GAB-017 demonstrated promising bioactivity in the subsequent in vitro assays ( Table 1 , Fig . 4 and Fig . 5 ) . GAB-001 exhibited selective binding to GAB1 ( Ki = 9 . 4±1 . 6 µM ) ( S6 Fig . ) , but not to IRS1 PH domain . In addition , it inhibited Y627 phosphorylation and killed breast cancer cells at IC50 = 23 . 7±2 . 7 µM ( for T47D ) . GAB-004 achieved similar binding selectivity as GAB-001 , but had a stronger inhibition of pY627 ( 81% ) and a lower IC50 ( 19 . 4±2 . 9 µM ) . Interestingly GAB-007 demonstrated weak binding ( KD = 42 . 3±8 . 9 µM ) and mild pY627 inhibition ( 53% ) , but it showed high cytotoxicity in both MDA-MB-231 ( IC50 = 4 . 6±0 . 7 µM ) and T47D ( IC50 = 20 . 1±4 . 3 µM ) cell lines ( Fig . 5 ) , probably due to other off-target mechanisms . GAB-016 and GAB-017 are N- ( 1 , 3 , 4-thiadiazol-2-yl ) benzenesulfonamide derivatives which were previously synthesized through our search for AKT PH domain inhibitors [28] . They demonstrated nanomolar binding affinity for GAB1 PH domain ( S6 Fig . ) , and were 5-fold and 28-fold more selective to the GAB1 than AKT , respectively . Consistent to their high binding affinities , GAB-016 and GAB-017 also inhibited over 80% of Y627 phosphorylation . All the aforementioned active inhibitors showed potent cytotoxicity to cancer cell lines ( T47D and MDA-MB-231 ) . More excitingly the cytotoxicity is specific to cancer cells as the inhibitors exhibit little inhibition in the non-cancer MCF-10A breast cell line ( Fig . 5 ) . Expectedly , as GAB1 and IRS1 pathways are intertwined [29] , some inhibitors could suppress IRS1 phosphorylation ( S7 Fig . ) . In addition , some compounds that selectively bind AKT PH domain ( e . g . , GAB-012 , GAB-013 and GAB-018 ) did not effectively kill MDA-MB-231 or T47D breast cancer cell lines at 50 µM ( Table 1 ) . To further investigate the structural mechanisms of our inhibitors to interact with the GAB1 PH domain , we performed MD simulations of the protein-inhibitor complexes ( listed in Table 2 ) . As expected , the active compounds ( GAB-001 , GAB-004 , GAB-007 , GAB-016 and GAB-017 ) demonstrated stable bindings to GAB1 PH domain in three independent simulations ( RMSD<2 . 5 Å ) , whereas GAB-002 and GAB-003 dissociated with the protein after around 25 ns ( S8 Fig . ) . In addition , MD simulations showed that GAB-007 , GAB-010 and GAB-016 could form stable binding to IRS1 PH domain ( S8 Fig . ) , consistent to the SPR results in Table 1 . To add another layer of validation of the binding modes predicted by MD simulations , we calculated the absolute binding free energies of our inhibitors to GAB1/IRS1 PH domain using an in-house potential of mean force ( PMF ) method , which aims to circumvent the insufficient sampling issue by introducing hypothetical intermediate states representing the association pathway of ligand from the unbound “bulk” regions to the ligand-binding “site” ( S9 Fig . ) . The principle of this approach has been described elsewhere [30] , [31] . Here , we implemented this method using ff99SB force field [32] . Briefly , the umbrella sampling and weighted histogram analysis were used as the primary tools to derive two sets of PMF: ligand conformational PMF w ( ξ ) and protein-ligand separation PMF w ( r ) . The details of mathematical calculations were available in S1 Method , and the w ( ξ ) and w ( r ) plots for eight protein-ligand complexes were available in S10 Fig . and S11 Fig . As indicated by Fig . 6 and Table 2 , the predicted absolute binding free energies via PMF method were in a good agreement with the experimental values ( RMSE = 0 . 64 kcal/mol , R2 = 0 . 85 ) . One may notice that these predictions encompassed two different PH domain targets ( GAB1 and IRS1 ) and a variety of ligand chemotypes . The good correlation between experimental binding free energies and predicted free energies implied that the predicted inhibitor binding modes by MD simulations were accurate . We have generated eight reliable PH domain-inhibitor complex models from MD simulations ( listed in Table 2 ) which have been validated by the PMF absolute binding free energy calculations as described in the previous section . When comparing the bound and unbound protein structures , we observed for the first time the ligand-induced conformational changes in three regions around the phoshpoinositide-binding pocket ( termed as Region I , Region II and Region III ) for both GAB1 and IRS1 PH domains . The Region I is comprised of the conserved K14GAB1/K21IRS1 ( β1[7] ) , K27GAB1/K21IRS1 ( β2[2] ) , Y47GAB1/Y46IRS1 ( β3[4] ) and F94GAB1/F93IRS1 ( β7[3] ) ( Fig . 7A–E ) . The MD simulations showed significant conformational changes in Region I ( RMSD>2 Å ) for both GAB1 and IRS1 , as illustrated by the RMSD analysis ( red plots in Fig . 7A ) . The side chain rearrangement of these residues , especially K14GAB1/K21IRS1 and Y47GAB1/Y46IRS1 , created a pocket which favorably binds an aromatic moiety connecting with a H-bond acceptor group ( Movie S1 , S2 , S3 available at http://imdlab . org/supporting/PLOSCompBio ) . This moiety could form cation-π and hydrophobic interactions with the surrounding K14GAB1/K21IRS1 and F94GAB1/F93IRS1 , respectively ( Fig . 7B–E ) . All inhibitors we identified in this study contain such a pharmacophore ( phenylthiazole in GAB-004 , phenylisoxazole in GAB-010 , S-phenyl carbothioate in GAB-007 , benzenesulfone in GAB-001 , GAB-016 and GAB-017 ) ( S12 Fig . ) . We had mentioned that β1[7] , β2[2] and β3[4] were the PIP3-binding residues , thus Region I conformational changes were attributable to the activities of the inhibitors . Generally , the conformational changes of Region I residues in GAB1 were more substantial than IRS1 , except F94GAB1 ( Fig . 7A ) . In comparison , GAB-010 could induce an alternative conformation of F93IRS1 ( Fig . 7E ) , which also occurred in ArhGAP9 crystal structure ( PDB ID: 2P0D [33] ) . The function of conformational change of F93IRS1 is likely to further open the pocket to accommodate larger moiety such as phenylisoxazole ( GAB-010 ) , as other IRS1 Region I residues were less flexible . The Region II is formed by β4 , β6 , 7 loop and the first several amino acids of β7 ( Fig . 7A–E ) . The key residues are R58GAB1/R62IRS1 ( β4[2] ) and R92GAB1/E91IRS1 ( β7[1] ) . Compared with Region I residues , more significant conformational changes were observed in the Region II residues in the GAB1 PH domain ( RMSD>2 . 5 Å ) ( blue plots in Fig . 7A ) . These conformational changes created a new pocket which binds aliphatic ( GAB-016 and GAB-017 ) and aromatic moieties ( e . g . , chlorobenzothiophene in GAB-001 and furan in GAB-004 ) . Remarkably , we found that the bulky aromatic moieties ( GAB-001 and GAB-004 ) generally induced more movement of GAB1 Region II residues than the aliphatic moieties ( GAB-016 and GAB-017 ) ( blue plots in Fig . 7A ) . We also observed that significant conformational changes only occurred in GAB1 , but not in IRS1 PH domain ( blue plots in Fig . 7A ) , probably because the electrostatic attraction between R62IRS1 and E91IRS1 significantly restrained the fluctuation of these two residues ( Movie S1 available at http://imdlab . org/supporting/PLOSCompBio ) , while the electrostatic repelling between R58GAB1 and R92GAB1 made these two residues more flexible ( Movie S3 ) . These findings imply that the flexibility of Region II residues of PH domain may correlate the size of binding group . The Region III is located on the solvent-accessible side of the β7 , especially I92GAB1 or H92IRS1 ( β7[3] ) ( Fig . 7A–E ) . When GAB-010 binds IRS1 PH domain , the benzimidazole moiety induced a significant side chain movement of H92 ( RMSD = 3 . 39 Å ) as compared with unbound form ( magenta plots on the right in Fig . 7A and Movie S3 available at http://imdlab . org/supporting/PLOSCompBio ) . In contrast , this region in GAB1 PH domain did not exhibit significant conformational changes when binding any inhibitor ( magenta plots on the left in Fig . 7A ) . Upon comparison of GAB1 and IRS1 PH domain sequences , we speculated that the accessibility of Region III was affected by the length of β1 , 2 loop: GAB1 PH domain has a longer β1 , 2 loop than IRS1 ( Fig . 2 ) , and the residues P16 and P17 forms intensive vdW interactions with β7 ( Fig . 2C ) , which would in turn block the access of inhibitors to Region III . This explains the selective binding of GAB-010 to IRS1 , but not GAB1 PH domain .
GAB1 is a critical protein in cellular signaling , and its PH domain has been suggested as an attractive target for various cancer treatments [34] , [35] . However , the absence of its 3D structure makes it challenging for structure-based drug discovery . We herein present a rigorously designed workflow for inhibitor identification by integrating various techniques ranging from structural bioinformatics , homology modeling , ligand-steered refinement , molecular dynamics , and virtual screening , followed by experiment evaluation with biochemical/biophysical and cellular assays . With our integrated protocol , we have successfully identified several selective inhibitors targeting the GAB1 PH domain and they are selective to breast cancer cells . This discovery offers us a great starting point to target this critical protein for cancer treatment , particularly for the triple negative breast cancer . Our results also showed that the triple-negative breast cancer cell line , MDA-MB-231 , was more resistant to GAB1 inhibitors than ER-positive breast cancer cell line , T47D ( Table 1 ) . It has been reported that MDA-MB-231 , but not T47D , has mutations on GAB1 downstream proteins , such as KRas and BRaf mutations [36] . Since KRas and BRaf mutations are known to reduce the dependency on the upstream activators , such as EGFR [37] , it was not surprising that MDA-MB-231 was more resistance to GAB1 inhibitors . Strikingly , we observed a concomitant inhibition of pGAB1 and pIRS1 by either GAB1-specific or IRS1-specific inhibitors ( S7 Fig . ) . This could be due to the crosstalk between c-Met and α6β4 integrin pathway [29] , which couples the phosphorylation of GAB1 and IRS1 upon HGF stimulation . These observations may bring us new insights of combined PH domain-targeted cancer therapeutic strategies . Further mechanistic studies are ongoing to investigate these hypotheses . While it is exciting to see that we have identified selective inhibitors of the GAB1 PH domain using our unique computation-experimentation integrated platform , we have to admit that some of the other hits also bind to multiple PH domains ( e . g . , IRS1 and AKT1 ) , as demonstrated by Table 1 . For example , GAB-001 and GAB-004 are selectively inhibitor GAB1 , but GAB-016 and GAB-017 are pan inhibitors against GAB1 , IRS1 and AKT1 . More follow-up experiments also showed that GAB-016 targets GAB2 PH domain as well . This is not surprising because PH domain is defined by their common β-sandwich structure . In addition , GAB1 and GAB2 PH domains are highly homologous ( 76% sequence identity ) , and IRS1 is one of the templates used in our homology modeling to build the 3D structure of GAB1 PH domain . Of note , all GAB1-selective or IRS1-selective inhibitors showed much better IC50 against T47D and MDA-MB-231 breast cancer cell lines than the non-tumorigenic MCF-10A cell line ( Fig . 5 ) . More intriguingly , we also observed that AKT1-selective inhibitors ( e . g . , GAB-012 and GAB-013 ) were toxic to MCF-10A at 100 µM , but not for T47D and MDA-MB-231 at the same concentration ( data not shown ) . This may imply that targeting GAB1 or IRS1 , but not AKT1 , might be a better targeted strategy for breast cancer treatment . Although PH domains have been intensively studied as cancer target for drug discovery , to date there are no available protein structures in complex with any drug-like small molecules . As mentioned , this has significantly limited the structure-based drug discovery efforts . In the present work , we utilized several inhibitors to investigate the dynamics of GAB1 PH domain and evaluate their selectivity in potential cancer cell inhibition . Interestingly , we found that the apo-structure of the PH domain protein could undergo large conformational changes in three regions to accommodate different inhibitors . The side-chain conformations of the residues in Region I determines the binding of either multiple electronegative groups ( e . g . , the multiple phosphates in IP4 ) or an aromatic moiety conjugated with a group containing H-bond acceptors ( e . g . , benzenesulfone ) , as shown in Fig . 7 . The accessibility of Region II and Region III depend on several critical amino acids on β4 and β7 and the length of β1 , 2 loop , respectively . The selectivity of PH domain inhibitors may be designed based on our modeling of the protein structures . For instance , GAB-010 is highly selective to IRS1 but no binding to GAB1 or AKT1 , largely due to the short β1 , 2 loop . The knowledge that GAB1 PH domain undergoes conformational change upon ligand binding provides novel insights of guiding the future structure-based drug design efforts , and of course more experimental validation will increase our understanding of GAB1 structure and functions .
A collection of five million drug and lead-like compounds which were curated from various sources ( e . g . , PubChem [38] and MayBridge ) was used for virtual screening . LigPrep [39] was employed for ligand preparation , including the removal of salts , assignment of appropriate protonation , tautomerization and ring conformations , and generation of 3D structures by energy minimization with OPLS2001 force field [40] . Additionally , an internal collection of 167 previously synthesized inhibitors targeting AKT PH domain were included for virtual screening . A total of 65 high-resolution crystal structures of PH domains were obtained from PDB , and we curated 34 non-redundant proteins . Their PDB IDs are listed in S2 Table . They were used for secondary structure-based sequence alignment with STRAP [15] . We extracted the multiple sequence alignments for β1 , β2 , β3 , β6 , β7 and α1 secondary structural fragments . The individual alignment will be used as input to PSI-BLAST [41] which could generate a PSSM for each individual fragment as shown in S3 Table . These PSSMs can be represented by WebLogo for more intuitive visualization and understanding ( S1 Fig . ) . These figures were generated using the WebLogo server [42] . The sequence of GAB1 PH domain was retrieved from UniProt database ( accession number Q13480 ) [43] . The secondary structure was predicted by PSSM combined with PSIPRED [16] and aligned to the templates ( myosin X ( PDB ID: 3TFM [17] ) , IRS1 ( PDB ID: 1QQG [18] ) , TAPP1 ( PDB ID: 1EAZ [19] ) and DAPP1 ( PDB ID: 1FAO [20] ) for homology modeling . To improve the quality of homology modeling , we manually corrected the multiple sequence alignment generated by ClustalX to ensure each secondary structure elements ( e . g . , α-helix and β-sheets ) were properly aligned . GAB1 PH domain homology models were built using MODELLER 9v10 [44] . As the active site residues of DAPP1 have the highest homology to those of GAB1 PH domain , the coordinates of the IP4 co-crystallized with DAPP1 was used as the initial structure . We generated initial 1 , 000 homology models . Since the lysine-rich loop β1 , 2 is important for phosphoinositide binding , especially for Group 1 PH domain [7] , [45] , the β1 , 2 loop of top ten initial models ( evaluated by DOPE score ) were subjected to ligand-steered refinement using built-in function of MODELLER . We selected five models out of the 100 generated loop models based on the overall DOPE scores [46] , Ramachandran plot , and the consistencies to IP4 binding site features [24] , [25] , [47]–[51] and the reported mutagenesis studies [6] , [8] . These five GAB1-IP4 complex models were refined by MD simulations using AMBER10 available at Texas Advanced Computing Center . All MD simulations were performed in triplicates with different initial velocities . The MD simulations were performed using ff99SB force field [32] in TIP3P explicit solvent with particle mesh Ewald ( PME ) , periodic boundary conditions and SHAKE . The topology and charges of the ligand were generated by Antechamber with AM1-BCC charges [52] . The system is solvated and neutralized in the cuboid box in which the closest distance between any atom originally in solute and the edge of the box is 12 Å . The system was equilibrated for 100 ps , and the production MD simulations were run in NPT ensemble for 20 ns , with the time step = 2 fs . The snapshots were taken every 1 ps . The root mean square deviation ( RMSD ) relative to the first frame and the root mean square fluctuation ( RMSF ) relative to the average structure were analyzed with cpptraj implemented in AmberTools12 [52] . The average structures were minimized , and the model quality was evaluated by QMEAN [21] , ProSA [22] and PROCHECK [23] . A reasonable protein model should have both ProSA and QMEAN Z-scores within the range for the native proteins of similar size , as illustrated by S3 Fig . GOLD 5 . 1 [53] was employed for virtual screening on our high performance computing cluster using the GAB1-IP4 complex model derived above . Molecular docking was performed with flexible side chains of the residues involved in IP4 binding , and the conformation with the best score of each compound was ranked based on their ChemPLP scores . Protein pharmacophore modeling was performed using GRID v22c [54] . Briefly , the GRID calculations were performed with a grid box enclosing the target with 1 Å beyond each dimension . During the calculations , the GRID directive Move was set ( MOVE = 1 ) to allow the flexibility of the side chains . The molecular interaction fields ( MIFs ) [55] were computed to determine the interaction between the receptor atoms and three different probes: the hydrophobic ( DRY ) , the amide nitrogen ( N1 , H bond donor ) , and the carbonyl oxygen ( O , H bond acceptor ) . Via visual inspection of the local minima of the GRID energy maps , the favorable binding sites of these three probes were used to define the features of a pharmacophore query . The derived pharmacophores were used to evaluate the binding poses of the initially selected 10 , 000 hits out of the five million compounds . If the docked hit poses fit the pharmacophore , they would be selected and subjected to clustering analysis based on the MACCS fingerprints and Tanimoto coefficient . The best scored compound from each cluster was chosen and the binding poses of these hits were individually inspected based on molecular visualization . In order to evaluate the selectivity of our inhibitors , we optimized the also PH domain ( IRS1 or GAB1 ) -inhibitor complex structures using MD simulations . The starting conformation for MD simulation is the binding mode which obtained the best score in molecular docking . The MD simulations were performed in triplicates for 50 ns using the parameters described in “3D structure prediction of GAB1 PH domain in complex with IP4” section . We also generated the trajectory of GAB1-GAB-001 complex ( Movie S1 ) , GAB1-GAB-017 ( Movie S2 ) and IRS1-GAB-010 ( Movie S3 ) . Each trajectory contained 1 , 000 snapshots which were taken every 50 ps . The ligands and the critical residues were in sticks , whereas the backbones of PH domain proteins were in ribbons . Starting from the docking conformation , these MD trajectories vividly demonstrated the conformational changes of the PH domain proteins upon ligand binding . The movies were available at http://imdlab . org/supporting/PLOSCompBio . The routine of PMF-based computation of protein-ligand absolute binding free energy has been previously described [30] , [31] . Briefly , the average structure of protein-ligand complex obtained from three independent 50 ns MD simulations was subject to energy minimization to remove clashes . The resulted structure was considered as the reference frame to define the position and orientation constraints . The PMF as a function of mass-weighted RMSD ( ξ ) relative to the reference ligand or the protein-ligand distance ( r ) was sampled by umbrella sampling and weighted-histogram analysis method ( WHAM ) . The full description of this method is available in S1 Method . The experimental binding free energies were derived from experimental KD ( or Ki ) using the equation or . The DNA sequences of human GAB1 and IRS1 PH domain ( IRS1 is for selectivity evaluation ) were cloned into pGEX-4T1 inducible bacterial expression plasmid ( GeneStorm , Invitrogen , Carlsbad , CA ) transformed into BL21 ( DE3 ) E . Coli . Expression and purification of the recombinant proteins were performed as previously described [51] . Binding assays were performed using a Biacore 2000 instrument with the Biacore Control Software v3 . 2 and BIAevaluation v4 . 1 analysis software ( Biacore , Piscataway , NJ ) as previously described [51] . Briefly , the PH domain GST-fusion proteins were immobilized on a CM5 Sensorchip ( Biacore BR-1000-12 ) using Biacore's Amine Coupling Kit ( Biacore BR-1000-50 ) to a level of 10 , 000 Response units ( RUs ) . Small molecule analytes at concentrations ranging from one tenth to ten times the predicted KD were injected at a high flow rate ( 30 µL/min ) . Dimethylsulfoxide ( DMSO ) concentrations in all samples and running buffer were 1% ( v/v ) or less . For the competitive binding assays and Ki determination , PtdIns ( 3 , 4 , 5 ) P3-biotin labeled liposomes ( Echelon Biosciences , Salt Lake City , UT ) and SA chips were used with increasing concentrations of the compound tested . We did triplicate SPR assays for each concentration . Two human breast cancer cell lines and one normal breast cell line were used for this study: T47D ductal breast epithelial tumor cell line , MDA-MB-231 epithelial tumor cell line and MCF-10A non-tumorigenic epithelial cell line ( American Type Culture Collection , Rockville , MD ) . T47D and MDA-MB-231 cells were maintained in bulk culture in Dulbecco's modified Eagle medium ( DMEM ) supplemented with 10% heat-inactivated fetal bovine serum ( FBS ) , 4 . 5 g/L glucose , 100 U/mL penicillin and 100 mg/mL streptomycin in a 5% CO2 atmosphere . MCF-10A cells were maintained in MEGM with other conditions same as the cancer cell lines . Cells were passaged using 0 . 25% trypsin and 0 . 02% EDTA . Cells were confirmed to be mycoplasma free by testing them with an ELISA kit ( Roche-Boehringer Mannheim , Indianapolis , IN ) . Our hit compounds were freshly prepared in DMSO at a stock concentration of 10 mM . For the evaluation of cellular proliferation , a standard 96-well micro-cytotoxicity assay was performed as described in reference [51] . Briefly , the assay was set up by plating cells at 5 , 000–10 , 000 cells per well ( depending on cell doubling time ) for a growth period of 4 days . The identified hits were added directly to the media , dissolved in DMSO at various concentrations ranging from 1 to 200 µM . The endpoint was spectrophotometric determination of the reduction of 3- ( 4 , 5-dimethylthiazol-2-yl ) -2 , 5-diphenyltetrazolium bromide . All assays were performed in triplicates . For all biological assays , hit compounds were added at 20 µM concentration directly into the culture media of the cells for 4 hr following a 16 hr incubation of T47D cells without FBS . Cells were stimulated with HGF for 20 min at 50 ng/ml . Following this treatment , cells were lysed as previously published [51] and equal amounts of total cell lysate were loaded on a pSer312-IRS-1/Total IRS-1 Meso Scale Discovery plate as described by the manufacturer . The plate was read using a Sector Imager 2400A instrument ( Meso Scale Discovery protein profiling system , Gaithersburg , MD ) . For the measurement of GAB1 phosphorylation , T47D cells were treated as for the phosphorylation of IRS1 evaluation . Cell lysates were run on a 7% SDS-PAGE and membrane were probed with specific anti-phospho-Tyr627 GAB1 ( Cell signaling ) . Each experiment was performed at least three times .
|
In this paper , we described the identification and evaluation of a set of first-in-class potent inhibitors targeting a new cancer target , Grb2-associated binder-1 ( GAB1 ) , which integrates signals from different signaling pathways , and is frequently over-expressed in cancer cells . To achieve our goals , we have employed intensive computational modeling to understand the structure of the GAB1 pleckstrin homology ( PH ) domain and screened five million compounds . Upon biological evaluation , we found that several inhibitors that induced large conformational changes of the target structure exhibited strong selective binding to GAB1 PH domain . Particularly , these inhibitors demonstrated potent and tumor-specific cytotoxicity in breast cancer cells . This represents a groundbreaking discovery in targeting GAB1 signaling which may be used for cancer therapy , especially for triple negative breast cancer patients .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biology",
"and",
"life",
"sciences",
"computational",
"biology"
] |
2015
|
Novel Inhibitors Induce Large Conformational Changes of GAB1 Pleckstrin Homology Domain and Kill Breast Cancer Cells
|
The leptospiral LigA protein consists of 13 bacterial immunoglobulin-like ( Big ) domains and is the only purified recombinant subunit vaccine that has been demonstrated to protect against lethal challenge by a clinical isolate of Leptospira interrogans in the hamster model of leptospirosis . We determined the minimum number and location of LigA domains required for immunoprotection . Immunization with domains 11 and 12 was found to be required but insufficient for protection . Inclusion of a third domain , either 10 or 13 , was required for 100% survival after intraperitoneal challenge with Leptospira interrogans serovar Copenhageni strain Fiocruz L1-130 . As in previous studies , survivors had renal colonization; here , we quantitated the leptospiral burden by qPCR to be 1 . 2×103 to 8×105 copies of leptospiral DNA per microgram of kidney DNA . Although renal histopathology in survivors revealed tubulointerstitial changes indicating an inflammatory response to the infection , blood chemistry analysis indicated that renal function was normal . These studies define the Big domains of LigA that account for its vaccine efficacy and highlight the need for additional strategies to achieve sterilizing immunity to protect the mammalian host from leptospiral infection and its consequences .
Pathogenic Leptospira species are globally distributed spirochetes that cause 350 , 000–500 , 000 severe human infections annually with an incidence of >10 cases per 100 , 000 population in humid , subtropical regions of the world and a mortality rate of 10% [1] , [2] , [3] . These figures are likely to be underestimates because leptospirosis is a neglected tropical disease that occurs more commonly among medically underserved populations [4] , [5] . The infection is endemic wherever there is exposure to urine of reservoir host animals that harbor the organism in their renal tubules [6] . At least 18 species and more than 200 leptospiral serovars have been described , many of which were isolated by cultivation of kidneys from a wide diversity of infested wild and domestic animals [1] , [7] . Environmental contamination of water and soil results in frequent outbreaks of leptospirosis among the poor in developing countries . Leptospirosis is also emerging among participants of aquatic sports and adventure tourism [8] , [9] . In the urban setting , Rattus norvegicus is the most important vector of human leptospirosis [5] . Serovars of Leptospira interrogans carried by rats cause life-threatening hepatorenal failure and pulmonary hemorrhage syndromes in tropical regions , especially where heavy rainfall occurs in urban areas with poor sanitation and flood control infrastructure [10] . Commercially available whole-cell bacterin vaccines for prevention of leptospirosis in animals provide relatively short-term serovar-specific protection and require frequent boosters [11] . Although inactivated whole-cell vaccines have been administered to humans , they are rarely used today because of their reactogenicity . Thus , there is an urgent need for development of novel vaccine strategies that provide safe , long-term , cross-protective immunity . Recombinant surface-exposed outer membrane proteins ( OMPs ) are attractive subunit vaccine candidates because in contrast to the lipopolysacchride , leptospiral OMPs are relatively well conserved and those that are surface-exposed represent potential targets for immune-mediated defense mechanisms . We have developed a suite of complementary approaches for determining which leptospiral OMPs are surface-exposed , including surface immunofluorescence , surface biotinylation , surface proteolysis , surface immunoprecipitation , and surface ELISA [12] , [13] , [14] , [15] . Using these approaches , a number of transmembrane OMPs and surface lipoproteins have been identified [16] , [17] . Despite the rapid increase in knowledge about leptospiral OMPs , progress in understanding their vaccine potential has been slow . Although LipL32 is the most abundant pathogenic leptospiral OMP [18] , purified , recombinant LipL32 has no detectable vaccine efficacy [19] . Nevertheless , hamsters immunized with recombinant bacillus Calmette-Guerin expressing LipL32 were partially protected from lethal challenge [20] and there is evidence for immunoprotection employing lipL32-containing viral or DNA-based vectors [21] , [22] . Synergistic immunoprotection has been observed using a combination of leptospiral OMPs , OmpL1 and lipidated LipL41 , expressed as membrane proteins in E . coli [23] . Leptospiral immunoglobulin-like ( Lig ) proteins are of great interest as mediators of leptospiral pathogenetic mechanisms , as serodiagnostic antigens , and as effective recombinant vaccinogens [24] , [25] , [26] , [27] , [28] . At least two of the three members of the Lig protein family are outer membrane lipoproteins containing a tandem series of bacterial immunoglobulin-like ( Big ) domains [29] . Lig protein expression is associated with virulence and is strongly and rapidly induced by increasing the osmolarity of the culture medium to physiologic levels found in the mammalian host , suggesting that they may be involved in the initial stages of host tissue colonization [30] , [31] . LigA consists of 13 Big domains , the first six of which are nearly identical in sequence to those in LigB , while the last seven are unique to LigA [32] and mediate interactions with host extracellular matrix proteins and fibrinogen [24] , [33] . One study has found that the region shared by LigA and LigB was not immunoprotective [27] , while another study reported that this region conferred some immunoprotective activity [34] . In contrast , several groups have reported that immunization with the LigA-unique region induced protection from lethal infection either in a mouse model [28] or in the hamster model [27] , [35] of leptospirosis . Although hamsters surviving leptospiral challenge were found to have sublethal kidney infection , both the extent of infection and its effects on the kidney , the key target organ in leptospirosis , were not well understood . In this study , we determined which LigA domains are most strongly associated with immunoprotection and the effect of LigA immunization on the burden of infection and the histopathology in the kidney . Our results show that protection from lethal infection required immunization with domains 11 and 12 along with a third domain , either 10 or 13 .
L . interrogans serovar Copenhageni strain Fiocruz L1-130 was maintained in Ellinghausen-McCullough-Johnson-Harris ( EMJH ) medium [36] supplemented with 1% rabbit serum ( Rockland Immunochemicals , Gilbertsville , PA ) and 100 µg/ml 5-fluorouracil at 30°C in a shaker incubator . Organisms were passaged no more than five times prior to hamster challenge . Hamster tissues were cultured in semi-solid EMJH or semi-solid Probumin Vaccine Grade Solution ( Millipore , Billirica , MA ) containing 0 . 2% Bacto agar ( BD , Franklin Lakes , NJ ) and 100 µg/ml 5-fluorouracil in a stationary incubator at 30°C and were examined for leptospiral growth for up to two months . PCR primers were designed to amplify gene fragments encoding various immunoglobulin-like domains from ligA of L . interrogans serovar Copenhageni strain Fiocruz L1-130 ( Table 1 ) . DNA amplicons , which included Nde I and Xho I restriction endonuclease sites , were ligated into pET-20b ( + ) ( Novagen ) , providing a carboxy-terminal His6 tag , and used to transform Escherichia coli BLR ( DE3 ) pLysS ( Novagen ) . Protein expression was induced with isopropyl-β-D-thiogalactopyranoside at 30°C and soluble proteins were released with BugBuster ( Novagen ) and purified with nickel-affinity chromatography as previously described [25] . All proteins were stored at 4°C after dialysis in PBS . Groups of four female Syrian hamsters , 5 to 6 weeks of age ( Harlan Bioscience , Indianapolis , IN ) , were immunized subcutaneously with 100 µg of recombinant protein , PBS , or 1×108 heat-killed ( 56°C for 1 h ) leptospires ( HKL ) in a total volume of 0 . 5 mL on days 0 , 14 and 28 with Freünd's adjuvant ( complete adjuvant for the first immunization , incomplete adjuvant for subsequent immunizations ) . Blood samples were obtained two days before the first immunization and 10 to 12 days after each immunization via the retro-orbital route . All animal procedures were approved by the Veterans Affairs Greater Los Angeles Healthcare System Institutional Animal Care and Use Committee and adhere to the United States Health Research Extension Act of 1985 ( Public Law 99–158 , November 20 , 1985 , “Animals in Research” ) , the National Institutes of Health's Plan for Use of Animals in Research ( Public Law 103–43 , June 10 , 1993 ) , U . S . Government Principles for the Utilization and Care of Veterbrate Animals Used in Testing , Research , and Training , Public Health Service Policy on Humane Care and Use of Laboratory Animals , the United States Department of Agriculture's Animal Welfare Act & Regulations , and Veterans Health Administration Handbook 1200 . 7 . Fourteen days after the third immunization ( day 42 ) , hamsters were challenged intraperitoneally with 1×103 L . interrogans serovar Copenhageni strain Fiocruz L1-130 in 0 . 5 mL of EMJH . The animals were weighed daily and observed for end-point criteria , including loss of appetite , gait or breathing difficulty , prostration , ruffled fur , or weight loss of ≥10% of the animal's maximum weight . Animals that reached end-point criteria were euthanized with isoflurane and tissue samples were collected in formalin for histopathology or incubated overnight at 4°C in RNAlater ( Ambion , Austin , TX ) and stored at −80°C . Processing tissues for histopathology involved formalin fixation , paraffin embedding , sectioning , and periodic acid Schiff ( PAS ) staining in a Dako automated slide processor . Blinded scoring of kidney sections used a scale of 0 to 5 for the extent of histopathology , ranging from normal to severe renal tubular damage , based on the degree of hyaline cast deposition , interstitial inflammation , mitosis , Bowman's space dilation , tubular atrophy and associated capsular depression . Blood was collected for serology and chemistry analysis ( Antech Diagnostics , Irvine , CA ) . 100 µL of blood or pulverized kidney or liver were inoculated into semi-solid medium at dilutions of 1:100 and 1:10 , 000 and incubated at 30°C . Sera collected at euthanasia were examined at a 1:50 dilution by MAT as previously described [38] with live L . interrogans serovar Copenhageni strain Fiocruz L1-130 . Briefly , heat-inactivated serum , diluted in physiologically buffered water , pH7 . 6 , was incubated overnight at 4°C with 2 to 4×108 leptospires/mL and examined under dark-field microscopy for >50% reduction in the number of free leptospires when compared with serum from uninfected animals . Kidneys were stored in RNAlater and DNA was extracted with DNeasy Blood and Tissue kit according to the manufacturer instructions ( Qiagen , Valencia , CA ) with modifications . 15 to 25 mg of kidney were immersed in 360 µL of ATL buffer and the tissue was homogenized in a 24-Fast Prep tissue homogenizer ( MP Biomedicals , Solon , OH ) using lysing matrix A with a setting of 6 m/s for 40s . 40 µL of proteinase K at a concentration of 15 mg/mL of protein were added and the samples were incubated for 3 h at 37°C . Two volumes of AL buffer-ethanol ( 1:1 ) were added and the mixture was applied to a spin column , on which the bound DNA was washed with washing solutions 1 and 2 and eluted with 200 µL of AE buffer-water ( 1:4 ) . The purified DNA was stored at −80°C until use . The extracted DNA was used in a qPCR using the Bio-Rad iQ5 Real-time System ( Bio-Rad , Hercules , CA ) . 100 ng of total DNA was combined with 1 µM of each primer and 12 . 5 µL iQ SYBR Green Supermix ( Bio-Rad ) and brought to a final volume of 25 µL with nuclease-free water ( Ambion ) . 4 samples were run per group and each sample was run in duplicate . qPCR primer pairs were LipL32-f , 5-CGCGTTACCAGGGCTGCCTT-3′ , and LipL32-r , 5′-CGCTTGTGGTGCTTTCGGTG-3′ , and hamster GAPDH-f , 5′-CTGGTTACCAGGGCTGCCTT-3′ , and GAPDH-r , 5′-CCGTTCTCAGCCTTGACTGTGC-3′ , resulting in amplicons of 152 bp and 146 bp , respectively . The PCR protocol consisted of an initial incubation step at 95°C for 12 . 5 min followed by 40 cycles of amplification ( 95°C for 15 s , 57°C for 30 s and 72°C for 30 s ) . The level of the lipL32 gene of L . interrogans was normalized to that of hamster gapdh , using Bio-Rad iQ5 software and Microsoft Excel . Standard curves were generated for each gene ranging from 10 to 1 . 6×106 copies of Leptospira ( 20-fold dilutions ) and 0 . 02 to 200 ng ( 10-fold dilutions ) of hamster DNA . Survival differences between groups were analyzed by Fisher's Exact Test using GraphPad InStat version 3 . 10 ( GraphPad Software Inc . , La Jolla , CA ) . One-way analysis of variance ( ANOVA ) was used to test for differences between multiple ( ≥3 ) groups using a P value<0 . 05 . For ordinal data , such as the histopathology scores , the Kruskal-Wallis one-way ANOVA with Dunn's post-test was included . The unpaired , two-tailed Student's t-test assuming unequal variance was used to test for differences between two groups using a P value<0 . 05 .
Eight clones were designed to express recombinant proteins corresponding to various LigA domains from the second half of domain 7 to domain 13 ( Table 1 ) of L . interrogans serovar Copenhageni . All proteins were expressed and purified as soluble proteins and found to be stable at 4°C after dialysis in sterile PBS . These proteins were employed as hamster immunogens in two independent experiments ( #1 and #2 ) and as antigens in an indirect ELISA to measure the corresponding antibody response . As shown in Figure 1 , hamsters had higher antibody titers after the third immunization than after one or two immunizations ( one-way ANOVA with test for linear trend , P<0 . 05 ) , except in the HKL ( experiment #2 ) and LigA7′-11 groups . There was no correlation between the antibody titer and the number of domains in the LigA protein ( Pearson correlation coefficient 0 . 29 , P>0 . 05 ) . Hamsters were challenged with virulent L . interrogans via the intraperitoneal route and observed daily , with a 10% decrease in body weight included as an end-point criterion . Body weight was found to be a useful measure of the response of animals to challenge; a decrease in body weight was the earliest observable sign of clinical leptospirosis . In contrast to animals that were immunized with LigA7′-13 and exhibited 100% challenge survival ( Figure 2A , Table 2 ) , non-surviving animals that were sham-immunized with PBS began to lose weight on day 8 after the challenge and reached -10% of peak weight within 48 h ( Figure 2B ) . Immunization with different recombinant LigA protein constructs ( Table 1 ) resulted in dramatically different challenge outcomes ( Table 2 and Figure 3 ) . In both experiments , there was 100% survival in hamsters immunized with either the LigA7′-13 or LigA10–13 proteins . In experiment #1 , immunization with either the LigA7′-11 protein or the LigA12-13 protein resulted in<50% survival . This result indicated that no single LigA domain was sufficient to afford 100% immunoprotection . For this reason , a second experiment was performed to identify the LigA domain ( s ) and the minimum number of domains required to protect hamsters from lethal challenge . Interestingly , both the LigA10-12 and the LigA11-13 proteins were both effective immunogens , while the LigA11-12 protein consisting of their shared domains afforded only 25% survival . Taken as a whole , these data indicate that LigA domains 11 and 12 are required but not sufficient to induce 100% survival . A recombinant LigA protein construct consisting of at least three specific Big domains is needed to induce a maximally protective immune response . The protective effect was not merely a reflection of antibody titer; as there was no correlation between survival and geometric mean end-point titer ( Figure 1 , one-way ANOVA , P > 0 . 05 ) . As previously reported [27] , immunization with LigA proteins provided non-sterilizing immunity , as organisms were isolated from the kidneys of animals surviving challenge . Cultures of kidney tissue from all hamsters surviving to 28 days were positive ( Table 2 ) . In contrast , only 3 and 10 of 56 animals had positive liver and blood cultures , respectively ( data not shown ) . One non-surviving animal immunized with LigA11-12 had a positive blood culture but negative cultures of the kidney and liver . The residual kidney infection was reflected in lower weight gain of hamsters after challenge ( Figure 4 ) . Among the surviving hamsters , those immunized with LigA10-13 had a non-statistical trend of gaining less weight after challenge than those immunized with LigA7′-13 or heat-killed leptospires . Infection resulted in the formation of agglutinating antibodies; the MAT was positive in nearly all LigA-immunized animals surviving for 28 days ( Table 2 ) . The only exceptions were one animal from the HKL control group and two from the LigA12-13 group that met end-point criteria early , the latter presumably because these animals had insufficient time to develop agglutinating antibodies . To more accurately assess the leptospiral burden , DNA from kidneys was analyzed by qPCR . As shown in Table 2 and Figure 5 , groups immunized with LigA fragments had a mean of 1 . 2×103 to 8×105 copies of leptospiral DNA per microgram of kidney DNA . As expected , kidneys from animals immunized with heat-killed leptospires had a lower leptospiral burden than groups immunized with LigA proteins such as LigA10-12 , LigA11-13 , LigA10-13 ( experiment #1 ) and LigA7′-13 ( experiment #1 ) ( non-parametric ANOVA , Dunńs post-test , P<0 . 05 ) . Leptospiral burden appeared to have a significant effect on animal health as reflected in the weight of surviving hamsters; there was an inverse correlation ( Pearson correlation coefficient -0 . 51 , P<0 . 05 ) in experiment #2 between the percent weight gain during the last week of the experiment and the copies of leptospiral DNA per μg of kidney tissue DNA . However , there were no significant differences in the leptospiral burden among groups with 100% survival immunized with different LigA proteins ( Non-parametric ANOVA , P>0 . 05 ) . Hemorrhagic areas were frequently noted on gross examination of the kidney and lungs of animals that did not survive challenge . Organs of survivors were usually normal in appearance but the kidneys occasionally appeared shrunken , pale , or had surface depressions indicating underlying infarction . Histopathological changes in the kidneys were largely limited to tubulointerstitial damage . Glomeruli were uniformly unaffected , except for one case of hyaline deposition seen in an HKL-immunized hamster . Although Bowman's space was dilated in some cases , the cells of the glomerulus were unaffected . Tubulointerstitial changes included renal tubular damage , encompassing changes of thinning of renal tubular epithelial cells ( compare Figures 6A and 6B ) , increasing hyaline cast deposition , mitosis , tubular atrophy ( Figure 6C ) , interstitial inflammation ( Figure 6D ) , and associated capsular retraction ( Figure 6E ) . Renal tubular obstruction was the most likely cause of hyaline cast deposition of the material staining intensely with PAS ( Figure 6F ) . Other changes due to tubular obstruction were dilated Bowman's space with or without hyaline casts . Mitoses were seen in only 2 cases , which further supported tubular injury because the rate of tubular cell turnover is normally close to zero . As shown in Table 2 , scores based on the extent of renal tubular damage were higher in groups immunized with the LigA10-12 and LigA11-12 proteins , suggesting that immunization with these constructs was associated with relatively more histopathology than other LigA constructs . Groups immunized with HKL and the LigA7′-13 protein had lower renal histopathology scores ( Table 2 ) and there was an inverse correlation between renal histopathology score and weight gain ( Pearson correlation coefficient -0 . 75 , P<0 . 01 ) . There was also an inverse correlation between renal histopathology score and leptospiral burden ( Pearson correlation coefficient −0 . 84 , P<0 . 01 ) for animals with>1 . 5×104 copies of leptospiral DNA/µg of tissue DNA , suggesting that a more intense immune response ( reflected by interstitial nephritis ) may be partially effective at clearing residual infection . Serum chemistries were measured to evaluate liver and kidney function of the hamsters ( Table 3 ) . Alanine aminotransferase and alkaline phosphatase levels were moderately elevated in all groups , consistent with hepatitis and cholestasis , respectively . However , bilirubin levels were universally normal , indicating that hepatic cholestasis had not progressed to biliary obstruction . Blood urea nitrogen ( BUN ) levels were increased in all groups and extremely elevated in the PBS control , while creatinine was low in all groups and elevated in the PBS control group ( one-way ANOVA with Dunn's post test , P<0 . 05 ) , indicating that renal dysfunction and/or dehydration contributed to mortality in these animals . In contrast , serum creatinine and BUN levels were universally normal in survivors , indicating that the renal tubular damage observed by histopathology had not progressed to frank kidney failure .
In this study , we identified the LigA domains involved in protecting hamsters from lethal leptospiral infection . Intraperitoneal inoculation was performed with 1000 L . interrogans serovar Copenhageni strain Fiocruz L1-130 , resulting in a lethal infection in all control animals ( Table 2 , Figure 3 ) . This is the same challenge dose used in a previously successful LigA protection study and is estimated to be ∼20-fold over the LD50 for this strain [27] . We found that a LigA protein construct consisting of at least three Big domains is required for immunoprotection and that the 11th and 12th specific Big domains must be included in this construct . Given that the average pairwise sequence identity among LigA Big domains is only 37% [32] , the domains identified here are likely to be antigenically unique and contain unique immunoprotective epitopes . Compared to maximally protective proteins , less protective LigA proteins elicited similar antibody titers in hamsters ( Figure 1 ) , suggesting that protection was not solely due to the antigenicity of the respective LigA vaccine . The mechanism of LigA mediated immunoprotection has not been elucidated , but may involve the disruption of a key function of LigA in leptospiral pathogenesis and/or the enhancement of host defense mechanisms . One key function of LigA is to mediate binding of Leptospira to host molecules such as fibronectin and fibrinogen [24] . Fibronectin- and fibrinogen-binding activity is found within domains 7 through 13 of LigA , with the carboxy-proximal domains 10 to 13 being required for fibronectin binding ( unpublished study , H . A . Choy ) . Finer mapping of the LigA binding activities may give clues as to the possible immunoprotective mechanism . As noted previously , LigA immunization converts an otherwise lethal infection into a sublethal kidney infection [27] . The burden of infection and its effects on vaccinated hamsters , qPCR and a histopathology scoring system were included as quantitative outcome measures . To our knowledge , this is the first vaccine study to use qPCR to quantitate leptospiral burden in animals after challenge . The application of qPCR to leptospiral vaccine studies allows for the accurate determination of the leptospiral burden , especially in the kidney , where colonization can lead to kidney damage and/or urinary shedding of the pathogen . We found that the heat-killed leptospires may not confer sterilizing immunity . Although the kidneys from the immunized animals were culture negative , leptospiral DNA was detected by qPCR . Reverse transcription-qPCR studies are needed to determine whether the low levels of DNA in these kidneys represent viable spirochetes or are remnants of leptospires killed by the host immune system . Comparison of quantitation results among surviving hamsters shows that immunization with as few as three LigA domains did not result in significantly higher levels of renal colonization than immunization with longer constructs such as the seven-domain LigA7′-13 protein ( Figure 5 ) . However , immunization with LigA10-12 did lead to greater histopathology , indicating different protective effects of the LigA10-12 and LigA11-13 constructs ( Table 2 ) . Histopathology analysis of kidney sections was performed using PAS staining , which is useful for evaluating many different types of nephropathology , including the severity of tubulointerstitial damage in our study . PAS staining facilitated identification of proximal tubules by their carbohydrate-containing brush border , evaluation of tubular basement membrane changes , as well as tubular atrophy ( Figures 6A , B , and C ) . A striking finding of our study was the identification of intensely staining protein casts in the tubules of 32% of animals , both in solid and “bubbly” deposition patterns ( Figure 6E ) . These protein casts probably represent Tamm-Horsfall glycoprotein ( THP ) , also known as uromodulin or TAMM protein , a glycoprotein that is produced by renal tubular epithelial cells [39] . THP is the most abundant protein in mammalian urine and though its deposition , in and of itself , is not pathologic , the high frequency of THP deposition in our study , including one case with extensive tubular deposition that occurred in an animal that succumbed to acute leptospirosis , suggests that increased THP deposition is related to the pathogenesis of leptospiral renal pathology . These physiologic hyaline deposits are usually solid , but in our study all cases demonstrated both a solid and “bubbly” deposition pattern . This “bubbly” pattern appeared to be due to a pathological process rather than an artifact of fixation and/or embedding , but further studies are needed to confirm this conclusion . Insufficient information is currently available to understand how broadly LigA immunoprotection can be applied . Whereas ligB has been found in all pathogenic Leptospira species , ligA has been found in only L . interrogans and L . kirschneri [32] . L . interrogans serovar Lai is the only L . interrogans isolate found not to contain ligA [40] . If ligA deficiency is confirmed in other Lai isolates , this would be a notable exception because the organism is both highly virulent and epidemiologically important . Recently , it was reported that homologous immunization with LigA7-13 that was expressed and purified under denaturing conditions did not protect hamsters against lethal infection by L . interrogans serovar Manilae strain L495 , an organism that expresses LigA [19] . This result stands in stark contrast to previously successful immunization studies involving L . interrogans serovars Manilae ( strain UP-MMC-NIID ) , Copenhageni and Pomona [27] , [28] , [41] . Although there were differences in the strains and adjuvants used , the finding that denatured LigA did not protect against lethal challenge could indicate that the protective epitope is conformational rather than linear . Accordingly , our finding that protective segments include domains 11 and 12 plus a third domain ( 10 or 13 ) on either end , suggests that three domains may be required for proper conformational folding . Additional research is needed to further define the structural requirements for LigA vaccine efficacy . We strongly recommend daily weighing of animals in leptospiral challenge experiments , including studies evaluating vaccine efficacy . We found that 10% weight loss effectively identified animals with leptospiral infection that had advanced to a premorbid condition . A similar result was observed in a recent study of leptospirosis in guinea pigs [42] . Weight loss is an objective end-point criterion that avoids uncertainty about whether an animal is able to eat and drink sufficient amounts of food and water . Thus , weight should be monitored along with other clinical parameters as different challenge doses or different strains may not present the same pattern of disease . In summary , we have mapped the immunoprotective segment of LigA and determined the minimal number of domains necessary to protect hamsters from lethal infection . This work also extends previous studies by quantifying the sublethal burden of infection and by defining the renal histopathological consequences of infection . It is worth noting that the immunoprotective domains we identified are contained within a segment that is known to mediate interactions with host extracellular matrix proteins [24] . This suggests that LigA-mediated immunoprotection may involve interference with key leptospiral-host interactions rather than a bactericidal mechanism . Further studies to define the kinetics of leptospiral infection in immunized animals may provide insight into both the mechanism of LigA-mediated immunoprotection and the development of vaccines for sterilizing immunity against leptospirosis .
|
Leptospirosis is the most widespread bacterial infection transmitted to humans from host animals that harbor the bacteria in their kidneys . Human infections caused by the bacterium , Leptospira interrogans , frequently result in a life-threatening illness characterized by jaundice and kidney failure . Vaccines are urgently needed to prevent leptospirosis in populations at risk . The leptospiral protein , LigA , is a promising vaccine candidate because it is the first purified protein to be shown to protect animals from fatal leptospirosis . The goal of this study was to determine which of LigA's 13 domains are required for the protective effect . Immunization with domains 11 and 12 was found to be required , but was insufficient , for protection . A third domain , either 10 or 13 , was required for 100% survival . As in previous studies , residual bacteria were cultured from the kidneys of survivors . However , in contrast to previous studies , we determined the amount of bacterial DNA in the kidneys as a measure of vaccine efficacy . We also examined the kidneys microscopically for signs of damage and measured blood chemistries to assess kidney function . These are important steps towards developing vaccines that provide protection from kidney damage and infection .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"adaptive",
"immunity",
"immunity",
"biology",
"microbiology",
"bacterial",
"pathogens"
] |
2011
|
A LigA Three-Domain Region Protects Hamsters from Lethal Infection by Leptospira interrogans
|
To measure the activity of neurons using whole-brain activity imaging , precise detection of each neuron or its nucleus is required . In the head region of the nematode C . elegans , the neuronal cell bodies are distributed densely in three-dimensional ( 3D ) space . However , no existing computational methods of image analysis can separate them with sufficient accuracy . Here we propose a highly accurate segmentation method based on the curvatures of the iso-intensity surfaces . To obtain accurate positions of nuclei , we also developed a new procedure for least squares fitting with a Gaussian mixture model . Combining these methods enables accurate detection of densely distributed cell nuclei in a 3D space . The proposed method was implemented as a graphical user interface program that allows visualization and correction of the results of automatic detection . Additionally , the proposed method was applied to time-lapse 3D calcium imaging data , and most of the nuclei in the images were successfully tracked and measured .
The animal brain is the most complex information processing system in living organisms . To elucidate how real nervous systems perform computations is one of the fundamental goals of neuroscience and systems biology . The wiring information for neural circuits and visualization of their activity at cellular resolution are required for achieving this goal . Advances in microscopy techniques in recent years have enabled whole-brain activity imaging of small animals at cellular resolution [1–4] . The wiring information of all the neurons in the mouse brain can be obtained using recently developed brain-transparentization techniques [5–9] . Detection of neurons from microscopy images is necessary for optical measurements of neuronal activity or for obtaining wiring information . Because there are many neurons in the images , methods of automatic neuron detection , rather than manual selection of ROIs ( regions of interest ) , are required and several such methods have been proposed [10 , 11] . Detection of cells that are distributed in three-dimensional ( 3D ) space is also important in other fields of biology such as embryonic development studies [12–17] . In these methods , cell nuclei are often labeled by fluorescent probes and used as a marker of a cell . To identify nuclei in such images , the basic method is blob detection , which for example consists of local peak detection followed by watershed segmentation . If the cells are sparsely distributed , blob detection methods are powerful techniques for nucleus detection . However , if two or more cells are close to each other , the blobs are fused , and some cells will be overlooked . These false negatives may be trivial for the statistics of the cells but may strongly affect individual measurements such as those of neuronal activity . Overlooking some nuclei should be avoided when subsequent analyses assume that all the cells were detected , for example , when making wiring diagram of neurons or establishing a cell lineage in embryonic development . Therefore , correct detection of all nuclei from images without false negatives is a fundamental problem in the field of bio-image informatics . Although many efforts have been made to develop methods that avoid such false negatives , these methods seem to insufficiently overcome the problem . In the head region of Caenorhabditis elegans , for example , the neuronal nuclei are densely packed and existing methods produce many false negatives , as shown below . Actually , in the studies of whole-brain activity imaging of C . elegans reported so far , the local peak detection method that can overlook many nuclei was employed [3 , 18] , or the nuclei were manually detected [19 , 20] . Highly accurate automatic nucleus detection methods should be developed in order to improve the efficiency and accuracy of such image analysis . Here we propose a highly accurate automatic nucleus detection method for densely distributed cell nuclei in 3D space . The proposed method is based on newly developed clump splitting method suitable for 3D images and improves the detection of all nuclei in 3D images of neurons of nematodes . A combination of this approach with a Gaussian mixture fitting algorithm yields highly accurate locations of densely packed nuclei and enables automatic tracking and measuring of these nuclei . The performance of the proposed method is demonstrated by using various images of densely-packed head neurons of nematodes which was obtained by various types of microscopes .
In this study , we focused on the head neurons of the soil nematode C . elegans , which constitute the major neuronal ensemble of this animal [21] . All the neuronal nuclei in a worm of strain JN2100 were visualized by the red fluorescent protein mCherry . The head region of the worm was imaged by a confocal microscope , and we obtained 3D images of 12 animals ( Data 1 , Fig 1A ) . The shape of the nuclei was roughly ellipsoidal ( Fig 1B ) . The fluorescence intensity increased toward the centers of the nuclei ( Fig 1D ) . The typical half-radius of the nuclei was about 1 . 10 μm ( S1 Fig ) . The distance to the nearest neighboring nucleus was 4 . 30 ± 2 . 13 μm ( mean and standard deviation , S1 Fig ) , suggesting that the neurons are densely distributed in 3D space . The mean fluorescence intensities differed among neurons by one order of magnitude ( S1 Fig ) , making it difficult to detect a darker nucleus near a bright nucleus . We first applied conventional blob detection techniques to the 3D image ( Fig 1C–1E ) . Salt-and-pepper noise and background intensities were removed from the image . The image was smoothed to avoid over-segmentation ( Fig 1C and 1D ) . Local intensity peaks in the preprocessed image were detected and used as seeds for 3D seeded grayscale watershed segmentation . Each segmented region was regarded as a nucleus ( Fig 1E ) . We found that dark nuclei in high-density regions often escaped detection . If the dark nucleus was adjacent to a bright nucleus , the fluorescence of the bright nucleus overlapped that of the dark one , and the local intensity peak in the dark nucleus was masked ( Fig 1D ) . As a result , the seed for the dark nucleus was lost , and the dark nucleus fused with the bright nucleus ( Fig 1E ) . The rate of false-negative nuclei was 18 . 9% . In contrast , our proposed method successfully detected and segmented the dark nuclei ( Fig 1F ) . The shapes of the nuclei are roughly ellipsoidal , and the fluorescence intensity increased toward the centers of the nuclei , suggesting that the intensity of nuclei can be approximated by a mixture of trivariate Gaussian distributions . The intensities fk of the k-th Gaussian distribution gk at voxel position x∈R3 can be written as fk ( x ) =πkgk ( x|μk , Σk ) =πkexp ( −12 ( x−μk ) TΣk−1 ( x−μk ) ) , where μk and Σk are the mean vector and covariance matrix of gk , respectively , and πk is an intensity scaling factor . To explain the effect on the curvature , typical bright and dark nuclei were approximated by the Gaussian distribution and are shown in Fig 2 as iso-intensity contour lines ( Fig 2A , 2C and 2E ) and plots of the intensity along the cross section ( Fig 2B , 2D and 2F ) . When a bright nucleus was near a dark nucleus , the peak intensity of the dark nucleus merged with the tail of the fluorescence intensity distribution of the bright nucleus and no longer formed a peak . These false negatives can be avoided by using methods for dividing a close pair of objects , or clump splitting . Such methods have been developed for correct detection of objects in two-dimensional ( 2D ) images [22–26] . These methods focus on the concavity of the outline of a blob . The concavity was calculated based on one of or a combination of various measurements such as angle [25] , area [27] , curvature [26] , and distance measurements [24] of the outline . In these methods , after binarization of the image , concavity was obtained for each point on the outline . Then the concave points were determined as the local peaks of the concavity . After determination of concave points , a line connecting a pair of concave points is regarded as the boundary between the objects . When we regard the outermost contour line in Fig 2E as the outline of the fused blob ( Fig 2G ) , the conventional 2D clump splitting method can be easily applied and two concave points are detected from the fused blob ( Fig 2G , red circles ) . The blob was divided into two parts by a border line connecting the two points , and the dark nucleus was detected . In the ideal case in Fig 2E , we obtained necessary and sufficient number of concave points . In real images , however , we might obtain too many concave points because outlines often contain noise and are not smooth . However , the number of concave points to choose is unknown because it is hard to know how many nuclei are included in a blob in a real image . Further , it is not obvious how to find the correct combinations of concave points to be linked if a blob contains three or more objects . In addition , for 3D images , the concepts of border lines that connect two concave points cannot be naturally expanded to three dimensions , because now we need some extra processes such as connecting groups of concave points in order to form border surfaces . Even if we regard a 3D image as a stack of 2D images , it is hard to split objects fused in the z direction ( direction of the stacks ) [11 , 27] . Here we introduce a concept of areas of concavity instead of concave points ( i . e . local peak of concavity ) . Hereafter we use curvature as a measure of concavity and focus on areas of negative curvature for simplicity and clarity , but other measures such as angle , area , and distance from convex hull may be applicable . Furthermore we used the iso-intensity contour lines inside the object in addition to the outline of the object . Near the concave points in Fig 2E , the iso-intensity contour lines have negative curvature; i . e . , they curve in the direction of low intensities . Negative curvature may be a landmark of the border line because a single Gaussian distribution has positive curvature everywhere . Actually , the voxels at which an iso-intensity contour line has negative curvature were between two Gaussian distributions ( Fig 2E , area between the broken lines ) . Once these voxels are removed from the blob , detection of two nuclei should be straightforward . This approach is different from the classic clump splitting methods in two respects; focusing on area rather than local peak of concavity ( concave points ) , and using iso-intensity contour lines in addition to the outline . These differences eliminate the need for determining how many concave points should be chosen and for obtaining correct combinations of the concave points because the area of negative curvature will cover the border lines . Therefore we can use the approach even if a blob contains three or more objects . In addition , this approach is robust to noise because it does not depend on a single contour line . Furthermore , this approach can be expanded to 3D images naturally because the 3D area ( i . e . voxels ) of negative curvatures will cover the border surfaces of the 3D objects . Iso-intensity contour lines in 2D images are parts of iso-intensity contour surfaces in three dimensions . A point on an iso-intensity surface has two principal curvatures , which can be calculated from the intensities of surrounding voxels ( S2 Text ) [28] . The smaller of the two principal curvatures is positive at any point in a single Gaussian distribution but is negative around the border of two Gaussian distributions . Therefore , once voxels that have negative curvature are removed from the blob , two or more nuclei should be detected easily in 3D images . Thus our approach solves the above problems of the classic clump splitting methods . We applied the above approach to real 3D images ( Fig 3 ) . The original images were processed by denoising , background removal , and smoothing to obtain the preprocessed images . The peak detection algorithm could find only a peak from the bright nucleus , and the blob obtained by watershed segmentation contained both nuclei . The principal curvatures of the iso-intensity surface were calculated from the preprocessed image . There were voxels of negative curvature in the area between two nuclei , but the area did not divide the two nuclei completely . The voxels of negative curvature were removed from the blob , and the blob was distance-transformed; these procedures were followed by 3D watershed segmentation . Thus , the two nuclei were separated , and the dark nucleus was successfully detected . After voxels of negative curvature were removed from the blobs , the size of blobs obtained by the second watershed segmentation tended to be smaller than real nuclei , and the distances between the blobs tended to be larger . To obtain the precise positions and sizes of the nuclei , least squares fitting with a Gaussian mixture was applied to the entire 3D image using a newly developed method ( see Methods ) . The number of Gaussian distributions and the initial values of the centers of the distributions were derived from the above results . Repeated application of watershed segmentation may increase over-segmentation . If the distance between two fitted Gaussian distributions is too small , the two distributions may represent the same nuclei . In this case , one of the two distributions was removed to avoid over-segmentation , and the fitting procedure was repeated with a single Gaussian distribution . The proposed method detected 194 out of 198 nuclei in the 3D image ( Fig 4 ) . Among the four overlooked nuclei , the intensities of two of them were too low to be detected . The other two had moderate intensities but were adjacent to brighter nuclei . In these cases , curvature-based clump splitting successfully split the two nuclei . However , deviations of the brighter nuclei from Gaussian distributions disrupted the fitting of the Gaussian distributions and resulted in misplacement of the Gaussian distributions for the darker nuclei , which were instead fitted to the brighter nuclei . On the other hand , the proposed method returned 11 false positives . Two of them resulted from the misplacement of the Gaussian distribution for the darker nuclei described above . Four of them were not neuronal nuclei but were fluorescence foci intrinsic to the gut . Three of them were the result of over-segmentation of complex-shaped non-neuronal nuclei . One of them was mislocalized fluorescence in the cytosol . The last one was the result of over-segmentation of a large nucleus that was fitted with two Gaussian distributions separated by a distance larger than the cutoff distance . We compared the performance of the proposed method with five previously published methods for nucleus segmentation ( Fig 5 and Table 1 ) . Ilastik [29] is based on machine learning techniques and uses image features such as Laplacian of Gaussian . FARSight [30] is based on graph cut techniques . RPHC [1] was designed for multi-object tracking problems such as whole-brain activity imaging of C . elegans and uses a numerical optimization-based peak detection technique for object detection . 3D watershed plugin in ImageJ [31] consists of local peak detection and seeded watershed . This method is almost the same as the conventional blob detection method used in our proposed method . CellSegmentation3D [32] uses gradient flow tracking techniques and was developed for clump splitting . This method has been used in the study of automated nucleus detection and annotation in 3D images of adult C . elegans [33] . We applied these six methods to 12 animals in Data 1 ( Fig 5 ) and obtained the performance indices ( Table 1 , see Methods ) . The parameters of each method were optimized for the dataset . The 3D images in the dataset contains 190 . 92 nuclei on average , based on manual counting . The proposed method found 96 . 9% of the nuclei and the false negative rate was 3 . 1% , whereas the false negative rate of the other methods were 11 . 2% or more . The false positive rate of the proposed method was 4 . 9% and that of the other methods ranged from 2 . 1% to 21 . 2% . The proposed method shows the best performance with both of the well-established indices , F-measure [12] and Accuracy [34] , because of the very low false negative rate and modest false positive rate . It should be noted that all of the compared methods overlooked more than 10% of nuclei in our dataset . The reason for this was suggested by the segmentation results , in which almost all of these methods failed to detect the dark nuclei near the bright nuclei and fused them ( Fig 5 , right column ) . These results suggest that all the compared methods have difficulty in handling 3D images with either large variance of object intensity or dense packing of objects , or both ( S1 Fig ) . These results clearly indicate that our proposed method detects densely distributed cell nuclei in 3D space with highest accuracy . Very low false negative rate is the most significant improvement of the proposed method from the other methods , suggesting that the proposed method will improve efficiency and accuracy of image analysis steps drastically . Because none of the computational image analysis methods is perfect , experimenters should be able to correct any errors they find . Therefore , a user-friendly graphical user interface ( GUI ) for visualization and correction of the results is required . We developed a GUI called RoiEdit3D for visualizing the result of the proposed method and correcting it manually ( S2 Fig ) . Because RoiEdit3D is based on ImageJ/Fiji [35 , 36] in MATLAB through Miji [37] , experimenters can use the familiar interface and tools of ImageJ directly . Developers can extend the functionality using a favorite framework chosen from various options such as ImageJ macros , Java , MATLAB scripts , and C++ languages . Interface with downstream analyses should be straightforward because the corrected results are saved in the standard MATLAB data format and can be exported to Microsoft Excel . Three-dimensional images are shown as trihedral figures using the customized Orthogonal View plugin in ImageJ ( S2 Fig ) . Fitted Gaussian distributions are shown as ellipsoidal regions of interest ( ROIs ) in each view . The parameters of the Gaussian distributions are shown in the Customized ROI Manager window in tabular form . The Customized ROI Manager and trihedral figures are linked , and selected ROIs are highlighted in both windows . When the parameters of the distributions or the names of nuclei are changed in the Customized ROI Manager window , the corresponding ROIs in the trihedral figures are updated immediately . Least squares fitting with a Gaussian mixture can be applied after ROIs are manually removed or added . RoiEdit3D can be used for multi-object tracking . The fitted Gaussian mixture at a time point is used as an initial value for the mixture at the next time point , and a fitting procedure is executed ( Fig 6A ) . Additionally , the intensities of nuclei can be obtained as parameters of the fitted Gaussian distributions . We tried to track and measure the fluorescence intensity of nuclei in real time-lapse 3D images ( Data2 ) . The animal in the image expressed a calcium indicator , so neural activity during stimulation with the sensory stimulus , sodium chloride , could be measured as changes in the fluorescent intensity . The proposed nucleus detection method was applied to the first time point in the image and found 194 nuclei out of 198 nuclei . Seventeen false positives and four false negatives were corrected manually using RoiEdit3D . Then the nuclei in the time-lapse 3D image were tracked by the proposed method . Most of the nuclei were successfully tracked . One or more tracking errors occurred in 27 nuclei during 591 frames , and the success rate was 86 . 4% , which is comparable to that in the previous work [1] . The tracking process takes 19 . 83 sec per frame ( total 3 . 25 hr ) . The ASER gustatory neuron was successfully identified and tracked in the time-lapse 3D image by the proposed method ( Fig 6B ) . The ASER neuron reportedly responds to changes in the sodium chloride concentration [38 , 39] . We identified a similar response of the ASER neuron using the proposed method ( Fig 6C ) . This result indicates that the proposed method can be used for multi-object tracking and measuring , which is an essential function for whole-brain activity imaging . Furthermore the proposed method was utilized to measure the fluorescence intensity of nuclei in time-lapse 2D images ( Data 3 ) . The proposed nucleus detection method was applied to the image for the first time point ( S3 Fig ) . Data 3 does not contain images of a highly-localized nuclear marker , and therefore the images of calcium indicator that was weakly localized to the nuclei were used instead . The proposed method found 7 nuclei out of 9 nuclei . Six false positives and two false negatives were corrected manually using RoiEdit3D . Then the nuclei were tracked by the proposed method . All of the nuclei were successfully tracked during 241 time frames . The ASER neuron was successfully identified and tracked in the 2D images . The response of the ASER neuron in the 2D images ( S3 Fig ) is similar to that in the 3D images . This result indicates that the proposed method can be used for multi-object tracking and measuring of 2D images as well as 3D images .
In this article , we proposed a method that accurately detects neuronal nuclei densely distributed in 3D space . Our GUI enables visualization and manual correction of the results of automatic detection of nuclei from 3D images as well as 2D images . Additionally , our GUI successfully tracked and measured multiple objects in time-lapse 2D and 3D images . Thus , the proposed method can be used as a comprehensive tool for analysis of neuronal activity , including whole-brain activity imaging . Although the microscopy methods for whole-brain activity imaging of C . elegans have been intensively developed in recent years [3 , 18–20] , computational image analysis methods were underdeveloped . In these works , the neuronal nuclei in the whole-brain activity imaging data were detected either manually or automatically by peak detection . Manual detection is most reliable but time- and labor-consuming , whereas the accuracy of the automatic peak detection is relatively low because of overlooking dark nuclei near bright nuclei . Our proposed method will reduce the difficulty and improve the accuracy . Furthermore , the numbers of the neuronal nuclei found or tracked in these four works were less than the real number of neuronal nuclei [3 , 18–21] . The scarcity may be due not only to the experimental limitations such as fluctuation of fluorescent protein expression or low image resolution , but also to the limitations of the image analysis methods that may overlook nuclei . The proposed method can detect almost all the nuclei in our whole-brain activity imaging data ( Fig 6 ) , suggesting that the proposed method can avoid errors that may be caused by overlooking nuclei , such as erroneous measurements of neural activities and misidentifications of neuron classes . Thus , our method will be highly useful for the purpose . Peng and colleagues have intensively developed the computational methods for automatic annotation of cell nuclei in C . elegans [33 , 40 , 41] . Although their methods successfully annotate cells in many tissue such as body wall muscles and intestine , the methods seem not to be applicable to annotations of head neurons in adult worms , which is highly desired in the field of whole-brain activity imaging [20] . They pointed out that the positions of neuronal nuclei in adult worms are highly variable [33] and this may be one of the reasons for the difficulty . The accuracy of detection and segmentation of neuronal nuclei may be another reasons because CellSegmentation3D that was incorporated in their latest annotation framework [33] shows compromised performance in our dataset ( Table 1 , Fig 5 ) . Our proposed method improves the accuracy of neuronal nucleus detection and will promote developing the automatic annotation methods for the neurons . It is noteworthy that the method of simultaneous detection and annotation of cells [41] is unique and useful in the studies of C . elegans . Because the method assigns the positions of reference to the sample image directly and avoid the detection step , the method find cells without overlooking under some conditions , but would not work correctly under the large variation of the numbers or the relative positions of the nuclei , both of which are observed in our dataset . The optimal method for accurate detection of nuclei will vary depending on the characteristics of the nuclei . Many conditions such as the visualization method , shape , and distribution of nuclei will affect these characteristics . In our case , the distributions of the fluorescence intensity of nuclei were similar to Gaussian distributions; thus , we developed an optimal method for such cases . Even if an original image does not have these characteristics , some preprocessing steps such as applying a Gaussian smoothing filter may enable application of our method to the image . Although choosing the optimal method and tuning its parameters might be more work than manual identification , the automatic detection method would improve subjectivity and effectivity . In the field of biology , it is often the case that hundreds or thousands of animals should be analyzed equally well . In such case , manual detection would be time-consuming and the automatic detection method would be required . For tracking the nuclei in time lapse images , we can apply the detection method to each time frame separately and then link the detected nuclei between frames . In this case , some false negatives and false positives would be separately produced for each frame , and they might disrupt the link step , resulting in increase of tracking errors . On the other hand , in the proposed method , the result of the automatic detection could be corrected manually , resulting in decrease of tracking errors . The proposed tracking method is a simplistic approach . Combination with existing excellent tracking methods will likely improve tracking performance of the proposed method . Cell division and cell death did not occur in our data , but they are fundamental problems in the analysis of embryonic development . It may be important to improve our method if it is to be applied to these problems so that the method handles such phenomena appropriately .
C . elegans strains JN2100 and JN2101 were used in this study . Animals were raised on nematode growth medium at 20°C . E . coli strain OP50 was used as a food source . We used three datasets in this study . Data1 and 2 contain ~200 neuronal nuclei , and Data3 contains 9 nuclei . The positions of the centers of the nuclei were manually corrected by experimental specialists using the proposed GUI . The blobs of the nuclei were detected by the conventional method ( Steps 1 & 2 ) . Under-segmented blobs were detected and split in Step 3 . The precise positions and sizes of the nuclei were obtained in Step 4 . The names and parameter values of the filters used in the proposed method are shown in S1 Table . The performance of proposed method for cell detection was compared with five state-of-the-art methods: Ilastik , FARSight , RPHC , 3D watershed plugin in ImageJ , and CellSegmentation3D . Ilastik is machine learning-based method and required a training data that was created manually . The parameters of RPHC was the same as the literature [1] . The parameters of the other methods were optimized based on F-measure and accuracy . The parallel displacements of the raw 3D images of 12 animals in Data 1 was corrected , and the methods were applied to the images . Because FARSight crashed during processing , its command line version ( segment_nuclei . exe ) was used [50] . The input images for FARSight and CellSegmentation3D were converted to 8-bit images because they could not operate with 32-bit grayscale images . For CellSegmentation3D , because it could not operate with our whole 3D image , the input images were divided and processed separately . The comparison was performed and the processing time was measured on the same PC as that used for the proposed method . All the methods other than CellSegmentation3D might be able to utilize multi-threading . The centroids of the segmented regions obtained by each program were used as the representative points of the objects . For the proposed method , the means of the fitted Gaussian distributions ( μk ) were used as the representative points . The Euclid distances of the representative points and manually pointed Ground Truth were obtained . If a representative point was nearest-neighbor of a point of Ground Truth and vice versa , the object was regarded as a True Positive . If only the former condition was met , the Ground Truth was regarded as a False Negative . If only the latter condition was met , the object was regarded as a False Positive . We obtained the indices of the performance [12 , 34 , 50] as follows: Truepositiverate=TPGT , Falsepositiverate=FPGT , Falsenegativerate=FNGT , F−measure=2×TP2×TP+FN+FP , Accuracy=TPTP+FN+FP , where GT=TP+FN . GT , TP , FP and FN mean Ground Truth , True Positive , False Positive and False Negative , respectively .
|
To reach the ultimate goal of neuroscience to understanding how each neuron functions in the brain , whole-brain activity imaging techniques with single-cell resolution have been intensively developed . There are many neurons in the whole-brain images and manual detection of the neurons is very time-consuming . However , the neurons are often packed densely in the 3D space and existing automatic methods fail to correctly split the clumps . In fact , in previous reports of whole-brain activity imaging of C . elegans , the number of detected neurons were less than expected . Such scarcity may be a cause of measurement errors and misidentification of neuron classes . Here we developed a highly accurate automatic cell detection method for densely-packed cells . The proposed method successfully detected almost all neurons in whole-brain images of the nematode . Our method can be used to track multi-objects and enables automatic measurements of the neuronal activities from whole-brain activity imaging data . We also developed a visualization and correction tool that is helpful for experimenters . Additionally , the proposed method can be a fundamental technique for other applications such as making wiring diagram of neurons or establishing a cell lineage in embryonic development . Thus our framework supports effective and accurate bio-image analyses .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"invertebrates",
"fluorescence",
"imaging",
"engineering",
"and",
"technology",
"caenorhabditis",
"neuroscience",
"animals",
"animal",
"models",
"caenorhabditis",
"elegans",
"human",
"factors",
"engineering",
"model",
"organisms",
"neuroimaging",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"imaging",
"techniques",
"animal",
"cells",
"man-computer",
"interface",
"calcium",
"imaging",
"graphical",
"user",
"interface",
"cellular",
"neuroscience",
"cell",
"biology",
"computer",
"architecture",
"neurons",
"nematoda",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"image",
"analysis",
"organisms",
"user",
"interfaces"
] |
2016
|
Accurate Automatic Detection of Densely Distributed Cell Nuclei in 3D Space
|
Histone ubiquitinations are critical for the activation of the DNA damage response ( DDR ) . In particular , RNF168 and RING1B/BMI1 function in the DDR by ubiquitinating H2A/H2AX on Lys-13/15 and Lys-118/119 , respectively . However , it remains to be defined how the ubiquitin pathway engages chromatin to provide regulation of ubiquitin targeting of specific histone residues . Here we identify the nucleosome acid patch as a critical chromatin mediator of H2A/H2AX ubiquitination ( ub ) . The acidic patch is required for RNF168- and RING1B/BMI1-dependent H2A/H2AXub in vivo . The acidic patch functions within the nucleosome as nucleosomes containing a mutated acidic patch exhibit defective H2A/H2AXub by RNF168 and RING1B/BMI1 in vitro . Furthermore , direct perturbation of the nucleosome acidic patch in vivo by the expression of an engineered acidic patch interacting viral peptide , LANA , results in defective H2AXub and RNF168-dependent DNA damage responses including 53BP1 and BRCA1 recruitment to DNA damage . The acidic patch therefore is a critical nucleosome feature that may serve as a scaffold to integrate multiple ubiquitin signals on chromatin to compose selective ubiquitinations on histones for DNA damage signaling .
Eukaryotic DNA is bound by histone proteins and organized into chromatin , the true in vivo substrate of transcription , replication and DNA repair , processes that are important in preserving genome integrity . Chromatin structure and function are highly regulated by histone post-translational modifications ( PTMs ) [1] . Histones are modified on distinct amino acid residues by different PTMs , such as phosphorylation , acetylation and ubiquitination , including several that are involved in DSB repair [2] . Upon DSB formation , H2AX is phosphorylated on Ser-139 within its C-terminal tail by the PIKK family kinases ATM , ATR and DNA-PK , to yield γH2AX [3] . γH2AX can be generated over a megabase of chromatin surrounding DSBs , thus creating microscopically-visible ionizing radiation-induced nuclear foci ( IRIF ) [4] , [5] . γH2AX creates a binding site for the DNA damage protein MDC1 , which promotes the localization of other DNA damage factors to damage sites [2] . Numerous E3 ubiquitin ligases including RNF8 , RNF168 , BRCA1 , RING1B and BMI1 are recruited to DNA lesions [6] , [7] . Collectively these DNA damage factors orchestrate the DNA damage response ( DDR ) that is a complex signaling network that is critical in regulating DNA damage signaling and repair [6] , [8] , [9] . Ubiquitin-mediated responses to DNA damage include histone H2A and variant H2AX ubiquitinations ( H2A/H2AXub ) . Indeed , H2A/H2AX is ubiquitinated by RNF168 , which targets Lys-13/15 within the N-terminal tail [10]–[12] , and RING1B/BMI1 that ubiquitinates C-terminal Lys-118/119 of H2A/H2AX [13]–[16] . Ubiquitinated histones H2AX and H2A mediate the chromatin association of both the mediator protein 53BP1 and the repair factor BRCA1 . These interactions occur through binding to Ubiquitin-interaction motif ( UIM ) domains in 53BP1 and in the BRCA1-interacting protein RAP80 [17] , [18] . Thus , site-specific histone ubiquitinations mediate critical signaling events that promote sensing and repair of DNA damage in mammalian cells [2] , [6] , [19] . Although the role of histone ubiquitination is well established in DNA damage signaling , it is unclear how the ubiquitin E3 ligases recognize their specific lysine targets on histones within the context of the nucleosome . Whether the nucleosome itself is involved in mediating the site-specific ubiquitin modifications on histones in response to DNA damage or other biological signals involving histone ubiquitinations has not yet been established . In this study , we find that the nucleosome acidic patch is required for RNF168- and RING1B/BMI1-dependent H2A and H2AX ubiquitination .
Ubiquitination of histones has emerged as a critical component of the DNA damage signaling pathway in mammalian cells [6] . We previously identified several mutations that reduced H2AX ubiquitin levels in undamaged cells [20] . One such mutation , H2AX-E92A resided in the acidic patch region of the nucleosome . Expression of tagged versions of human H2AX and H2A in human HEK293T cells revealed a full-length protein species of predicted size as well as a slower migrating ubiquitinated form for both human H2AX and H2A ( Figure S1A , B ) . Mutation of glutamic acid 92 to alanine ( E92A ) reduced H2AX and H2A ubiquitination ( H2AX/H2Aub , Figure S1A , B ) . These results identify the amino acid E92 of human H2AX/H2A as an important residue for H2AX/H2Aub . We next sought to define the contribution of the acidic patch region of the nucleosome towards H2AX/H2Aub and the DDR . H2AX/H2A is specifically ubiquitinated on the N-terminal Lys-13/15 by RNF168 [10]–[12] , as well as on the C-terminal Lys-118/119 by RING1B/BMI1 [13]–[16] . Therefore , an important question was to determine which sites on H2AX rely on the acidic patch for ubiquitination . To answer this question , we first created a lysine-free human H2AX where all lysine residues were mutated to arginines . These mutations maintain the basic charge at each amino acid location but are unable to be ubiquitinated ( Figure 1A ) . As expected , expression of H2AX-allR in HEK293T cells confirmed that this mutant lacked any detectable ubiquitination , similarly to H2AX-E92A ( Figure 1B ) . Unlike these H2AX derivatives , mutation of the DNA damage induced phosphorylation site on H2AX ( S139 ) to an unphosphorylatable residue ( S139A ) did not affect H2AXub ( Figure 1B ) . Having identified a mutant H2AX derivative that lacked ubiquitination , we then reverted specific arginine residues in this mutant back to lysine residues that are contained in WT H2AX ( Figure 1A ) . This strategy allowed us to unambiguously identify site-specific ubiquitinations within H2AX . As expected , H2AX mutants lacking K118/119 exhibited a large reduction in mono-ubiquitination ( Figure 1C ) . This confirmed previous work showing that these sites on H2AX/H2A are the major lysine acceptor sites for mono-ubiquitination [2] , [21] . Interestingly , we observed ubiquitination of the H2AX derivative containing only K13/15 as acceptor sites for ubiquitin ( Figure 1C ) . We also observed an increase in K13/15ub on this H2AX derivative upon DNA damage , which is consistent with previous studies showing that a small fraction of H2AX becomes ubiquitinated on K13/K15 following DNA damage by the E3 ubiquitin ligase RNF168 [10]–[12] . To assess the contribution of the acidic patch towards H2AX K13/15ub , we tested whether an E92A mutation would affect H2AX K13/15ub . Combining the E92A mutation within the H2AX derivative that could only be ubiquitinated on K13/15 abolished any detectable ubiquitination at these sites within H2AX ( Figure 1C ) . We next tested whether the acidic patch also affected the ubiquitination of K118/119 of H2AX . Analysis of an H2AX derivative that could only be ubiquitinated on K118/119 showed that this protein was readily ubiquitinated and mutation of the acidic patch diminished H2AXub at these specific lysine sites ( Figure S2 ) . RNF168 is a limiting factor within the DDR and overexpression of RNF168 increases H2AX-K13/15ub levels but not H2AX-K118/119ub levels [10]–[12] , [22] . In agreement with these studies , we observed that overexpression of RNF168 increased H2AX-K13/15ub but not H2AX-K118/119ub ( Figure 1D , S2 ) . In accordance with our results from Figure 1C , mutation of the acidic patch decreased H2AX-K13/15ub levels , even under conditions where RNF168 is overexpressed and not limiting ( Figure 1D ) . H2AX-K13/15ub is mediated by RNF168 whose recruitment to sites of DNA damage requires MDC1 and RNF8 [23] , [24] , which in turn require H2AX phosphorylation on S139 [24]–[27] . Collectively , these findings suggest that γH2AX may be required for H2AX-K13/15ub . To test this possibility , we mutated S139 within the H2AX-allR-R13/15K derivative to monitor specifically H2AX-K13/15ub in either the presence or absence of S139 . While E92A abolished H2AX-K13/15ub , the S139A mutation did not affect ubiquitination at these sites ( Figure 1E ) . These results show that S139 is not required in cis for H2AX-K13/15ub under these conditions . We note that these experiments were performed in cells containing WT H2AX that could provide functional residues in trans for H2AX-K13/15ub . These experiments were done in the presence of overexpressed RNF168 , which could bypass the requirement for S139 phosphorylation for its recruitment to chromatin . In overexpression conditions , RNF168 accumulates at sites of endogenous DNA damage marked by 53BP1 [22] , which we note requires K13/15ub on H2A/H2AX [10] . Regardless , under either limiting or non-limiting conditions for RNF168 , we find that the acidic patch is required for H2AX-K13/15ub ( Figure 1C–E ) . Our results strongly suggested that the acidic patch is required for both K13/15 and K118/119 H2AX/H2Aub . We next sought to test whether the effect of the acidic patch mutation on H2AX/H2Aub was direct , as well as to analyze the role of the acidic patch in mediating site-specific ubiquitinations with their associated E3 ligases . To assess these questions , we reconstituted H2AX and H2A nucleosome core particles ( NCPs , Figure 2A–C ) with or without the acidic patch mutation ( i . e . E92A ) and subjected them to in vitro ubiquitination ( Ub ) assays . Previous studies have established that bacterially expressed and purified RNF168 and RING1B/BMI1 complexes catalyze the specific addition of ubiquitin on H2AX/H2A NCPs at K13/15 and K118/119 respectively [10] , [11] . Using the same constructs and experimental conditions , we performed in vitro Ub assays with H2AX NCPs with or without the E92A acidic patch mutation . As expected , RING1B/BMI1 and RNF168 ubiquitinated H2AX within WT NCPs ( Figure 2D , E ) . In contrast , both E3 ligase complexes were unable to efficiently ubiquitinate NCPs containing E92A mutation in H2AX ( Figure 2D , E ) . These effects appear to occur within the context of the nucleosome as RNF168 could ubiquitinate the free form of H2AX whether it was WT or contained the acidic patch E92A mutation ( Figure 2F ) . We performed identical experiments with H2A WT and E92A NCPs and obtained the same results ( Figure 2G–I ) . As another control , we subjected H2AX and H2A WT and E92A NCPs to in vitro methylation assays with SET8 , a methyltransferase that is active only within the context of the nucleosome for methylating H4K20 [28] . The acidic patch mutation did not affect nucleosome specific SET8 methylation suggesting the E92A mutation does not overtly disorder the NCP ( Figure S3 ) . These in vitro results are consistent with our in vivo data and demonstrate that RNF168 and RING1B/BMI1 require the nucleosome acidic patch of H2AX/H2A to promote site-specific ubiquitination of H2AX/H2A K13/15 and H2AX/H2A K118/119 respectively . Our findings show that the nucleosome acidic patch mediates both H2AX/H2A K13/15ub by RNF168 and H2AX/H2A K118/119ub by RING1B/BMI1 . Several studies have shown that RING1B/BMI1 participates in the DDR although a clear function for H2AX/H2A K118/119ub is as yet unidentified [13]–[16] , [29]–[31] . In contrast , the function of H2AX/H2A K13/15ub by RNF168 was recently elucidated and is well defined [10] . Indeed , RNF168-dependent H2AX/H2A K15ub is selectively recognized by the ubiquitination-dependent recruitment motif ( UDR ) of 53BP1 that , together with its Tudor domain , reads a bivalent ubiquitin-methylation signal at DNA damage sites to recruit the DDR factor 53BP1 . A clear prediction of this mechanism is that 53BP1 recruitment to sites of DNA damage would be perturbed in the absence of H2AX-K13/15ub and/or H2A K13/15ub . We chose to next focus on the role of the acidic patch in regards to RNF168-dependent H2AX/H2A K13/15ub in vivo since we could utilize 53BP1 foci formation as an in vivo read-out for functional H2AX/H2A-K13/15ub . Because RNF168 specifically targets H2AX-K13/15 , we sought to characterize further our H2AX derivatives where K13/15 are the only lysines available for ubiquitination and to ascertain the contribution of both the acidic patch and RNF168 expression levels on H2AX-K13/15 ubiquitin levels . Expression of SFB-tagged WT H2AX resulted in clearly identifiable mono-ubiquitinated species whose electrophoretic mobility was retarded as expected due to the presence of a 9 kDa ubiquitin protein ( Figure 3A ) . Rendering WT H2AX unmodifiable by ubiquitin on all but K13/15 resulted in an almost complete loss of mono-ubiquitinated H2AX ( Figure 3A ) . This reduction was also observed when the acidic mutation E92A was added to this H2AX derivative . To analyze the contribution of both RNF168 and the acidic patch on H2AX-K13/15ub , we repeated these experiments in the presence of overexpressed Myc-tagged RNF168 . Although we still observed reduced H2AXub in K13/15 only H2AX derivatives compared to WT H2AX , we now were able to specifically detect H2AX-K13/15ub using this H2AX derivative that only contained K13/15 ( Figure 3A ) . Interestingly , we were able to detect a small increase in H2AX-K13/15ub upon DNA damage suggesting that this H2AX derivative was functioning within the DDR in cells ( Figure 3A ) . Under these optimized conditions for specifically detecting H2AX-K13/15ub , mutation of the acidic patch ( i . e . E92A ) resulted in a large reduction in H2AX-K13/15ub levels either in the presence or absence of DNA damage ( Figure 3A ) . Thus , we could detect DNA damaged induced H2AX-K13/15ub in the presence of RNF168 , and in all conditions tested , H2AX-K13/15ub required the acidic patch . Having now characterized H2AX derivatives for their ubiquitination on K13/15 , K118/119 or in the absence of lysines , we sought to determine whether H2AX ubiquitinations were required in vivo for the DDR and more specifically for 53BP1 foci formation . Up to now , all of our experiments analyzing H2AX derivatives were performed in the presence of WT H2AX . To overcome this limitation , we turned to a human cell line deleted for H2AX , MCF10A H2AX−/− , that we previously characterized [20] . To test the contribution of H2AXub for 53BP1 IRIF , we stably reconstituted MCF10A H2AX−/− with WT H2AX and derivatives to compare the ability of site-specific mutations in ubiquitinated sites on H2AX to complement the defect of 53BP1 IRIF that occurs in these cells in the absence of H2AX . We first created stable cell lines expressing H2AX constructs to be tested and selected clones for each that expressed H2AX in the majority of cells and to similar protein levels as the WT H2AX reconstituted cell line ( Figure 3B , S4 and data not shown ) . To assess 53BP1 IRIF , we analyzed several H2AX derivatives for their ability to rescue defective 53BP1 IRIF in MCF10A cells lacking H2AX . As we previously reported , MCF10A H2AX−/− and MCF10A H2AX−/−+H2AX S139A are unable to support equivalent recruitment of 53BP1 into IRIF compared to WT MCF10A cells ( Figure 3C ) . Surprisingly , all H2AX derivatives tested , including a lysine-less H2AX ( allR ) that cannot support ubiquitination on either K13/15 or K118/119 , were able to fully support 53BP1 IRIF ( Figure 3C ) . Thus , although S139 phosphorylation is required for 53BP1 IRIF in these cells , H2AXub ( including K13/15 or K118/119 ) , as well as the H2AX acidic patch , is dispensable for 53BP1 IRIF ( Figure 3C ) . As DNA damage dependent H2A-K13/15ub also occurs , these results suggest that γH2AX could function in trans to promote H2A- K13/15ub that would be sufficient to mediate 53BP1 recruitment to sites of DNA damage . One hypothesis could be that DNA damage induced H2AX phosphorylation on S139 could mediate an initial ubiquitination on H2AX-K13/15 that would be required to amplify RNF168-dependent H2Aub . Similarly , the nucleosome acidic patch of H2AX could initiate the recruitment and activation of RNF168 that would in turn trigger the start of this ubiquitin-dependent signaling pathway . However , our results argue against these hypotheses and instead suggest that the acidic patch of H2A , as well as H2Aub , can compensate for H2AXub in the DDR to support 53BP1 IRIF . Testing the role of the H2A acidic patch and H2Aub in vivo is extremely challenging due to the unavailability of a mutation system for H2A in human cells . Regardless , our findings establish that the acidic patch of H2AX , as well as H2AXub , is dispensable for 53BP1 IRIF in human cells . To overcome the limitations of studying histone mutants in vivo and to validate the requirement of the nucleosome acidic patch in promoting H2AX/H2Aub and subsequent DDR signaling , we sought to identify an experimental approach to target the acidic patch regions of both H2A and H2AX in vivo . The nucleosome acidic patch of H2A has been shown to interact with several proteins including histone H4 , the Kaposi's sarcoma–associated herpesvirus ( KSHV ) protein LANA , IL-33 , HMGN2 and RCC1 [32]–[36] . The finding that several proteins interact through this nucleosome region has suggested that the nucleosome acidic patch acts as a “chromatin platform” to mediate various cellular signals via their interactions with chromatin through the acidic patch . As our data has identified the nucleosome acidic patch of H2AX and H2A as a requirement for RNF168- and RING1B/BMI1-dependent H2AX/H2Aub in vitro and in vivo , we set out to test whether expression of a known acidic patch interacting protein could interfere with these DDR factors . This experimental approach has the advantage of blocking both H2A and H2AX acidic patch regions , a potential necessity for uncovering the function of this nucleosome domain in the DDR . Results from these experiments would further define the role of the nucleosome acidic patch of both H2A and H2AX in the DDR and would allow us to test our hypothesis that the acidic patch of H2A and H2AX functions in the DDR in vivo , at least in part by promoting H2AX/H2A-K13/15ub . The KSHV latency-associated nuclear antigen ( LANA ) interacts with the nucleosome acidic patch of H2A to tether episomes to chromosomes [32] . The first 32 amino acids of LANA comprise the acidic patch interacting region and expression of a GFP fusion with this minimal region in cells is sufficient to target this small truncated region of the protein to mitotic chromosomes [32] . Additionally , mutation of the 8–10 amino acid region ( named 8LRS10 ) of this 32 amino acid LANA peptide abolishes the interaction of LANA with the nucleosome acidic patch . To assess whether this acidic patch interacting peptide from LANA could compete with RNF168- and RING1B/BMI1-dependent H2AX/H2Aub , we synthesized the minimal acidic patch interacting peptide from LANA along with the 8LRS10 mutant peptide and analyzed the effects of these peptides on our previously characterized in vitro Ub assays . Interestingly , the acidic patch binding LANA peptide reduced H2Aub that was catalyzed by both RING1B/BMI1 and RNF168 in a concentration-dependent manner ( Figure 4A , S5 ) . The reduction of H2Aub by the LANA peptide required the ability to bind the acidic patch as the 8LRS10 mutant peptide was unable to compete away H2Aub . These results supported our previous findings that the acidic patch was directly promoting histone ubiquitination by these E3 ligases and also suggested that the LANA peptide could interfere with this reaction in cells . To begin to address this question , we wanted to ask whether we could observe a decrease in H2AXub in cells expressing LANA peptide . We cloned and engineered a GFP-fusion of LANA containing only the first 32 amino acids ( GFP-LANA ( 1–32a . a . ) , [32] ) . Next , we co-transfected our H2AX derivatives with GFP-LANA and analyzed H2AXub by western blotting . We observed that the ubiquitination of WT H2AX , H2AX-K118/119 only and H2AX-K13/15 only were reduced when co-expressed with GFP-LANA in cells ( Figure 4B–D ) . These results are in agreement with both our in vitro and in vivo data demonstrating that the nucleosome acidic patch of H2AX is required for K13/15 and K118/119 ubiquitination ( Figure 1 , 2 ) . The ability of LANA to inhibit H2AXub in vivo suggested that cells expressing LANA would exhibit impaired DNA damage signaling . If this were indeed the case , a clear prediction would be that cells expressing LANA would exhibit reduced 53BP1 IRIF due to H2AX/H2Aub inhibition from LANA blocking RNF168 through the acidic patch . To test this possibility , we expressed GFP-LANA in human U2OS and HEK293T cancer cells and analyzed 53BP1 IRIF with and without GFP-LANA . Upon DNA damage , we observed reduced 53BP1 IRIF in cells expressing GFP-LANA compared to GFP alone expressing cells ( Figure 5A , B , S6 ) . Importantly , the upstream DDR factor MDC1 , as well as γH2AX , were unaffected by GFP-LANA expression ( Figure 5A–C ) . This is consistent with RNF168 inhibition by LANA as RNF168 acts downstream of γH2AX and MDC1 [23] , [37] . To rule out any potential cell cycle effects due to GFP-LANA expression , we analyzed the cell cycle of GFP-LANA expressing cells . Analysis of these cells using FACS , DNA labeling by hoechst and phospho-Histone H3 ( S10 ) immunostaining , a histone mark specific for mitotic cells , did not reveal any detectable differences in cell cycle stage or DNA staining between control and GFP-LANA expressing cells ( Figure S7A–D ) . In addition , expression of mutant GFP-LANA-8LRS10 , a mutation that is unable to bind the acidic patch , had no discernable effect on 53BP1 IRIF showing that the effect of GFP-LANA on the DDR required its interaction with the nucleosome acidic patch ( Figure 5C–D ) . We also confirmed the inhibition of 53BP1 , but not MDC1 , in GFP-LANA expressing cells by laser micro-irradiation ( Figure 5E ) . We observed that cells expressing high levels of GFP-LANA were able to fully inhibit 53BP1 recruitment to laser damage compared to cells expressing lower levels of GFP-LANA ( Figure 5E ) . These results are consistent with GFP-LANA targeting the nucleosome acidic patch resulting in inhibition of 53BP1 recruitment to DNA damage . 53BP1 functions in DNA double-strand break repair by both promoting NHEJ and inhibiting HR ( reviewed in [38] ) . 53BP1 recruits the DDR factor RIF1 to DNA damage sites where it inhibits DNA end-resection and acts as the main effector of 53BP1-dependent NHEJ [39]–[43] . Consistent with GFP-LANA inhibiting RNF168-dependent 53BP1 recruitment , we also observed reduced RIF1 accumulation at IRIF in GFP-LANA expressing cells ( Figure 6A ) . RNF168 is also required for the recruitment of the HR factor BRCA1 to DNA damage sites [23] , [37] . Interestingly , GFP-LANA also impaired BRCA1 IRIF in S/G2 cells ( Figure 6B; S/G2 cells were identified by CyclinA positive staining ) . Quantification of IRIF in GFP-LANA expressing cells revealed a greater than 50% reduction in cells with greater than 10 foci for either RIF1 or BRCA1 ( Figure 6C , D ) . The ability of GFP-LANA to impair IRIF of DDR factors appears to be dependent on expression levels . We observed that high LANA expressing cells displayed a greater reduction in DDR factor recruitment compared to low LANA expressing cells , which explains the incomplete inhibition of DDR factor recruitment to DNA damage sites by GFP-LANA ( Figure 5E , 6B ) . 53BP1 also inhibits DNA-end resection in G1 to block HR and promote NHEJ [39] , [41] , [42] , [44] . Since expression of GFP-LANA impaired 53BP1 foci formation at DNA damage sites , we analyzed whether these cells exhibited functional inhibition of 53BP1 by monitoring DNA end-resection in G1 cells . RPA is recruited to , and binds , resected DNA , which is normally restricted to CyclinA-positive S/G2 cells . As expected , in control cells that do not express GFP-LANA or cells expressing mutant GFP-LANA-8LRS10 , RPA foci at laser damage were virtually undetectable using our experimental conditions ( Figure 6E , F , quantified in G ) . Interestingly , GFP-LANA expressing cells readily formed RPA foci at laser damage in CyclinA-negative G1 cells ( Figure 6E , F , quantified in G ) . Thus , GFP-LANA expression resulted in DNA end-resection in G1 cells , which supports our previous results showing impaired 53BP1 recruitment to DNA damage by GFP-LANA . Taken together , these results are consistent with a role for the nucleosome acidic patch in promoting both 53BP1 and BRCA1 DDR pathways by mediating RNF168-dependent DNA damage signaling in vivo . In summary , our results support a model whereby RNF168 and RING1B/BMI1 require the nucleosome acidic patch on H2AX/H2A to target these histones on site-specific lysines and that GFP-LANA can inhibit these processes ( Figure 7 ) . By overcoming the limitations of mutating the acidic patch of both H2A and H2AX through the expression of GFP-LANA , we have determined that the nucleosome acidic patch functions in vivo to promote RNF168-dependent DNA damage signaling . We have also created a novel tool that has the ability to silence DNA damage signaling at the level of RNF168 as well as inhibit RING1B/BMI1-dependent H2AX/H2Aub in vivo , which could be useful for studying these ubiquitin-dependent processes in cells . Of note , some viruses inactive the DDR by ubiquitin-dependent degradation mechanisms that target DDR factors , including RNF168 [45] , [46] . Our results suggest that viruses , including LANA expressing KSHV , could inactive the DDR through another means by interfering with the nucleosome acidic patch . This potential mechanism would inhibit H2A/H2AX ubiquitination and subsequent DNA damage responses whose inhibition can affect viral transcription and activation of latent viruses in mammalian cells [47] . Additionally , other nucleosome acidic patch binding factors , including RCC1 and HMGN2 , could also potentially affect the DDR . RCC1 and HMGN2 have opposing effects on chromatin dynamics with RCC1 promoting condensation of DNA prior to mitosis and HMGN2 decompacting chromatin through interactions with linker histone H1 [48] . We envision that these factors could regulate the DDR in multiple ways including chromatin dynamics and/or competition with other nucleosome acidic patch interacting proteins including RNF168 and RING1B/BMI1 . Additional studies are warranted to investigate the interplay between nucleosome interacting factors and the DDR . To our knowledge , this study has identified the first nucleosome domain that participates in both H2A/H2AXub and the DDR in human cells . Most studies have focused on the role of histone modifications , including ubiquitination , in the DDR . Our findings provide evidence that the DDR engages the nucleosome acidic patch , which participates in promoting histone ubiquitinations that mediate DDR factor interactions with chromatin including 53BP1 . Chromatin interaction motifs within both RNF168 and RING1B/BMI1 have been identified . For example , RNF168 contains multiple ubiquitin-binding domains that target RNF168 to chromatin [49] and the RING1B/BMI1 complex contains DNA binding activity that is critical for histone ubiquitination [50] . Similar to the bivalent reading of histone marks by 53BP1 , our results suggest that the histone ubiquitin writers , RNF168 and RING1B/BMI1 , utilize multivalent chromatin interactions , including the nucleosome acidic patch , to write their “histone code . ”
HEK293T and BOSC23 cells were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin . Human U2OS cells were grown in Dulbecco's modified Eagle's medium ( DMEM ) supplemented with 10% fetal bovine serum ( FBS ) , 100 U/ml penicillin , 100 µg/ml streptomycin and 2 mM L-glutamine . WT and H2AX-deficient MCF10A cells were cultured in DMEM/F12 medium supplemented with 5% horse serum , EGF ( 20 ng/ml ) , hydrocortisone ( 0 . 5 mg/ml ) , cholera toxin ( 100 ng/ml ) , insulin ( 10 µg/ml ) and 1% penicillin/streptomycin . Cells were kept at 37 C in a humidified incubator containing 5% CO2 . Human H2AX cDNA was a generous gift from Dr . Michael Huen from The University of Hong Kong . WT and mutant human H2AX cDNAs were cloned into gateway compatible entry vector ( pDONR201 ) . The cDNAs were then subcloned into expression vectors harboring N-terminal SFB ( S-protein/2×Flag/Streptavidin-binding peptide ) , 3×Flag , Myc epitope tag , GFP epitope tag or HA-Flag epitope tag as indicated . Bacterial expression vectors for core histones ( human H2A , H2B , H3 and H4 ) in pET21 were previously described [51] . Human H2AX cDNA were cloned into gateway compatible entry vector ( pDONR201 ) and subcloned into bacterial expression vector pDEST17 harboring N-terminal 6× His tag . Constructs containing the nucleosomal 601 sequence was a kind gift from Ilya Finkelstein ( UT Austin ) . 5′-biotin tagged 601 nucleotide sequence was generated by PCR using the primer pairs: 5′ ( Btn ) CTGGAGAATCCCGGTGCC ( forward primer ) and 5′ACAGGATGTATATATCTGACACG ( reverse primer ) to be used for reconstitution of nucleosomes . pET24b ( + ) -Bmi1-His6 ( residues 1–108 ) , pGEX-6P-1-RING1B ( residues 1–116 ) and pPROEX HTa-RNF168 ( residues1–113 ) were obtained as described [10] . Full length LANA cDNA was a gift from Chris Sullivan ( UT Austin ) . The N-terminal 32 amino acids of LANA were PCR amplified with the following primers Forward: 5′-TTGTCGACATGGCGCCCCCGGGAATGCGCCTGA-3′; Reverse: 5′-TTTCTAGACTATCTTTCCGGAGACCTGTTTCG-3′ and cloned into eGFP-C1 ( Clontech ) vector using 5′-SalI and 3′-XbaI restriction sites to create GFP-LANA ( 1–32a . a . ) . Primers for mutating 8RLS10 to AAA of LANA have been described by a previous study and were used to mutate GFP-LANA ( 1–32a . a . ) to GFP-LANA-8LRS10 [52] . All mutations were generated using site-directed mutagenesis following standard protocols . All plasmid inserts and mutations were confirmed by DNA sequencing . The primary antibodies used were as follows: mouse anti-FLAG antibody ( Sigma-Aldrich; F1804 ) , mouse anti-γH2AX ( Cell Signaling; #9718 ) and ( Millipore; #05-636 ) , rabbit anti-53BP1 ( Novus Biologicals; NB100-304 ) , mouse anti-53BP1 ( BD transduction laboratories; 612522 ) , rabbit anti-MDC1 ( Abcam; ab11169 ) , goat anti-Rif1 ( N-20 ) ( Santa Cruz , sc-55979 ) , rabbit anti-beta-tubulin ( Abcam; ab6046 ) , mouse anti-c-Myc ( Santa Cruz; sc-40 ) , rabbit anti-H2AX ( Cell Signaling; #2595 ) . Secondary antibodies for western blotting were as follows: anti-rabbit igG , HRP-linked ( Cell Signaling; #7074 ) , anti-mouse IgG , HRP-linked ( Cell Signaling; #7076 ) . Secondary antibodies for IF analysis from Invitrogen were as follows: Alexa Fluor 488 ( Rabbit , A11034 ) , Alexa Fluor 594 ( Rabbit , A11037 ) , Alexa Fluor 594 ( Mouse , A11032 ) , Alexa Fluor 594 ( Goat , A11058 ) , Alexa Fluor 647 ( Rabbit , A21245 ) and Alexa Fluor 647 ( Mouse , A21236 ) . For experiments involving cell cycle analysis , the following antibodies were used; mouse anti-BRCA1 ( D-9 ) ( Santa Cruz; sc-6954 ) , mouse anti-RPA32/RPA2 [9H8] ( Abcam; ab2175 ) , rabbit anti-Phospho-histone H3 ( Ser10 ) ( D2C8 ) ( Cell Signaling , #3377 ) and rabbit anti-CyclinA ( H-432 ) and ( Santa Cruz; sc-751 ) . Mammalian expression ( SFB- , Myc- and GFP- ) vectors were transfected using lipofectamine 2000 according to manufacturer's instruction and HA-Flag-retroviral expression constructs were co-transfected with pCL-ampho in BOSC23 cells using Lipofectamine 2000 ( Invitrogen ) according to the manufacturer's instruction . Viruses were harvested and filtered at 48 h and 72 h after transfection . MCF10A H2AX−/− cells were transduced by virus containing medium and selected by puromycin ( 2 µg/ml ) . The GFP-LANA ( 1–32a . a ) or GFP-LANA-8LRS10 constructs were transfected into the U2OS cells using HilyMax ( Dojindo ) according to the manufacturer's instruction . After 24 h post-transfection , cells were treated with 2 Gy IR and processed 2 h post-treatment . A Faxitron X-ray machine ( Faxitron X-ray Corporation ) was used for gamma irradiation ( IR ) . Mammalian cells were lysed in NETN ( 150 mM NaCl , 1 mM EDTA , 10 mM Tris-Cl , pH 8 . 0 , 0 , 5% Nonidet P-40 ( v/v ) containing protease inhibitors . Samples were separated by SDS-PAGE in sample loading buffer , transferred to PVDF membranes , incubated overnight in primary antibodies as indicated , followed by 1 h of incubation in HRP-conjugated secondary antibodies . Western blots were detected by standard chemiluminescence ( GE Healthcare Amersham ECL prime ) using a Bio-Rad Molecular Imager ChemiDoc XRS+ system . U2OS cells were plated on glass-bottomed dishes ( Willco Wells ) . Post 8 h of transfection with GFP-LANA ( 1–32a . a ) , cells were pre-sensitized with 10 µM of 5-bromo-2′deoxyuridine ( BrdU ) in normal DMEM medium for 20 h . Laser micro-irradiation was carried out with a Fluoview 1000 confocal microscope ( Olympus ) . Laser setting and protocols were as previous described [53] . After incubation with the indicated time points , cells were fixed and analyzed by immunofluorescence and microscopic imaging as described below . For quantification , >50 cells were scored for all conditions from at least two independent experiments . U2OS , HEK293T and MCF10A cells were grown on poly-L-lysine Cellware 12 mm round coverslips ( BD Biosciences ) . After the indicated treatments in HEK293T or U2OS cells , samples were treated and processed for IF as previously described [54] . For MCF10A cells , cells were pre-extracted by incubating coverslips in CSK buffer ( 10 mM PIPES , pH 6 . 8 , 100 mM NaCl , 300 mM sucrose , 3 mM MgCl2 , 1 mM EGTA , 0 . 5% ( v/v ) Triton X-100 ) for 10 min on ice before fixing followed by IF analysis as previously described [20] . Cells were imaged using an inverted Fluoview 1000 confocal microscope ( Olympus ) and Z-stacked images were analyzed with Fluoview 3 . 1 software . For IRIF quantification , >100 cells were counted for all conditions . Data was analyzed in Prism and graphs were plotted from data obtained from two or three independent experiments as indicated . U2OS cells were transfected with either GFP-LANA , Myc-LANA or a control vector for 8 h . 24 h after transfection , cells were harvested and fixed for 24 h with 80% ethanol . The fixed samples were then washed three times with 1× PBS containing 1% FCS and incubated with phospho-histone H3 ( S10 ) primary antibody for 2 h followed by incubation of goat anti-rabbit secondary antibody for 2 h at room temperature for the mitotic index assay . For DNA content analysis , the cells were washed and stained with propidium iodide and processed on a BD Accuri C6 flow cytometer ( BD Biosciences ) for cell cycle analysis .
|
Post-translational modifications of histones play important roles in regulating both the structure and function of chromatin . As all DNA based processes , including transcription , DNA replication and DNA repair , occur within the context of chromatin , the actual in vivo substrate of these reactions is chromatin . Thus , understanding these processes within the context of chromatin is vital for providing mechanistic insights into chromatin-based processes , including DNA damage signaling and genome maintenance . Here we identify a structure within H2A and H2AX termed the acidic patch that promotes the activity of two independent ubiquitin E3 ligase complexes , RNF168 and RING1B/BMI1 , and is required for DNA damage ubiquitin signaling . We show directly in vitro and in vivo that this nucleosome structure is critical for histone H2A and H2AX ubiquitinations and the DNA damage response in cells . In addition , we engineered a novel biological tool that blocked the nucleosome acidic patch of all histone H2A species leading to the repression of the DNA damage response in cells . Collectively , DNA damage factors elicit their response not only through histone modifications such as ubiquitin but also through interactions within nucleosome surface structures to activate DNA damage signaling .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] |
[
"cellular",
"stress",
"responses",
"genetic",
"mutation",
"chromosome",
"biology",
"gene",
"expression",
"genetics",
"biology",
"molecular",
"cell",
"biology",
"chromatin",
"histone",
"modification"
] |
2014
|
Nucleosome Acidic Patch Promotes RNF168- and RING1B/BMI1-Dependent H2AX and H2A Ubiquitination and DNA Damage Signaling
|
Tetanus is a vaccine-preventable , neglected disease that is life threatening if acquired and occurs most frequently in regions where vaccination coverage is incomplete . Challenges in vaccination coverage contribute to the occurrence of non-neonatal tetanus in sub-Saharan countries , with high case fatality rates . The current WHO recommendations for the management of tetanus include close patient monitoring , administration of immune globulin , sedation , analgesia , wound hygiene and airway support [1] . In response to these recommendations , our tertiary referral hospital in Tanzania implemented a standardized clinical protocol for care of patients with tetanus in 2006 and a subsequent modification in 2012 . In this study we aimed to assess the impact of the protocol on clinical care of tetanus patients and their outcomes . We examined provision of care and outcomes among all patients admitted with non-neonatal tetanus to the ICU at Bugando Medical Centre between 2001 and 2016 in this retrospective cohort study . We compared three groups: the pre-protocol group ( 2001–2005 ) , the Early protocol group ( 2006–2011 ) , and the Late protocol group ( 2012–2016 ) and determined associations with mortality by univariable logistic regression . We observed a significant increase in provision of care as per protocol between the Early and Late groups . Patients in the Late group had a significantly higher utilization of mechanical ventilation ( 69 . 9% vs 22 . 0% , p< 0 . 0001 ) , provision of surgical wound care ( 39 . 8% vs 20 . 3% , p = 0 . 011 ) , and performance of tracheostomies ( 36 . 8% vs 6 . 7% , <0 . 0001 ) than patients in the Early group . Despite the increased provision of care , we found no significant decrease in overall mortality in the Early versus the Late groups ( 55 . 4% versus 40 . 3% , p = 0 . 069 ) , or between the pre-protocol and post-protocol groups ( 60 . 7% versus 50 . 0% , p = 0 . 28 ) . There was also no difference in 7-day ICU mortality ( 30 . 1% versus 27 . 8% , p = 0 . 70 ) . Analysis of the causes of death revealed a decrease in deaths related to airway compromise ( 30 . 0% to 1 . 8% , p<0 . 001 ) but an increase in deaths due to presumed sepsis ( 15 . 0% to 44 . 6% , p = 0 . 018 ) . The overall mortality in patients suffering non-neonatal tetanus is high ( >40% ) . Institution of a standardized tetanus management protocol , in accordance with WHO recommendations , decreased immediate mortality related to primary causes of death after tetanus . However , this was offset by an increase in death due to later ICU complications such as sepsis . Our results illustrate the complexity in achieving mortality reduction even in illnesses thought to require few critical care interventions . Improving basic ICU care and strengthening vaccination programs to prevent tetanus altogether are essential components of efforts to decrease the mortality caused by this lethal , neglected disease .
Despite being a vaccine-preventable disease , tetanus is frequently encountered in sub-Saharan Africa [2 , 3] . The incidence of non-neonatal tetanus cases has fallen since the initiation of the vaccination programme but the number of cases remains high , with 4 , 604 non-neonatal cases reported in 2016 in the African region , [4] and likely many more that were not reported [5] . Among global tetanus deaths , 44% occur in sub-Saharan Africa and the highest proportion of these is in East Africa [6] . Inadequate vaccination is cited as the primary causative factor for tetanus despite the availability of a highly effective vaccine [7 , 8] . In Tanzania , national tetanus vaccine coverage is 87% in children less than 1 year , as determined by history and vaccination cards , but regional discrepancies are high [9] . In Mwanza , where our hospital is located , only 70% of infants received all basic vaccines in 2015 [10] . In addition , because the Tanzanian vaccination programme focuses on children under the age of 1 and pregnant women who attend antenatal clinic , there is currently no system in place to ensure booster vaccination for men past the infancy doses , despite recommendations by the WHO [11] . This would explain why young men are the most at risk of tetanus infection in Tanzania and many other sub-Saharan African countries [3] . A recent study from Tanzania showed that only 28% of men older than 15 years were seroprotected against tetanus [12] . This has tremendous economic and social consequences: in 2015 the WHO calculated the cost of a single dose of tetanus vaccine at $0 . 14 [13] whereas the cost of caring for patients with tetanus in low- and middle-income countries ranges from $78 [14] to $900 [15] . In Tanzania , 55% of the population lives in extreme poverty ( less than $1 . 25 per day ) and men are the main financial providers , so the socioeconomic impact is tremendous . Lack of medications , inadequate implementation of proven treatment interventions , high treatment cost for patients , and long distance to specialised centres have been cited as additional key reasons for high tetanus-associated mortality in sub-Saharan Africa [3 , 16 , 17] . Studies of targeted approaches to address these barriers are lacking . Therefore , our goal was to conduct a quality-improvement project to assess whether the utilization of a standardized hospital protocol for management of tetanus was effective in reducing mortality . Our study was possible because in 2006 our Tanzanian referral hospital implemented a hospital protocol to be used for management of all patients admitted to our ICU with tetanus , with further updates in 2012 . We sought to determine whether the implementation of this protocol had any impact on provision of clinical care and patient outcomes . We hypothesised that protocol-driven implementation of proven tetanus interventions would increase over the study period , and that tetanus mortality would decrease .
Ethical approval for the conduct of this study was obtained from the joint Catholic University of Health and Allied Sciences ( CUHAS ) /BMC Research Ethics and Review Committee ( BREC/001/18/2008 ) , the National Institute for Medical Research ( NIMR/HQ/R . 8c/Vol . IX/1085 ) , and Weill Cornell Medicine ( 1108010827 ) . We conducted a retrospective cohort study of all patients who had a diagnosis of tetanus and were admitted to the Intensive Care Unit ( ICU ) of Bugando Medical Centre ( BMC ) , Tanzania , from May 2001 to September 2016 . BMC is a public tertiary referral hospital located in Mwanza , a north-western city on the shores of Lake Victoria . Mwanza is the second largest city in Tanzania and BMC serves the 15 million people of the Lake Zone . The ICU admits approximately 500 patients annually from all disciplines , with specialists from internal medicine and anaesthesia providing the majority of the ICU care . The ICU has 13 beds with 7 mechanical ventilators , pressurised wall oxygen , suction , and bedside monitors . The monitors display non-invasive blood pressures , saturations and electrocardiography . There is regular availability of intravenous ( IV ) fluids , antibiotics , adrenaline , and dopamine . Central lines and noradrenaline are intermittently available . The nurse to patient ratio is 1:2 , with most of the nurses having no formal training in intensive care medicine . In view of high mortality rates and in an effort to improve the quality of care for patients , BMC implemented a hospital protocol to optimize care of tetanus patients in 2006 , in accordance with WHO recommendations . The hospital protocol can be seen in the supporting information ( Supporting information ( S1 ) ; Fig 1 . Bugando Medical Centre tetanus management protocol ) In 2012 , several additional key interventions were implemented in the ICU , including employment of a dedicated ICU physician and an agreement with the surgical department that tracheostomies would be performed as soon as possible for tetanus patients admitted to the ICU . The stepwise protocol is organised based on clinical urgency of interventions . The initial stage focuses on airway management either in the form of oral intubation or immediate tracheostomy . There was no specific criterion or threshold for initiation of mechanical ventilation during the entire study period and this decision was made at the discretion of the on-duty physician . This is followed by emphasis on early prescription and administration of immune globulin , antibiotics , spasm control with magnesium sulphate and benzodiazepines with accompanying analgesia , fluid management to avoid acute kidney injury and deep vein thrombosis prophylaxis . Wound care is also included in the early stages with surgical consultation for wound debridement if needed . Farmers and manual workers were identified , from previous work , as the key at-risk group , and most present with identifiable wounds . The next stage focuses on monitoring the patient’s response to interventions and conducting investigations . Instructions are provided on how to adjust prescribed medication to get the desired outcomes . The final stage focuses on the recovery phase of care with instructions of down-titrating the medications and ensuring immunisation prior to ICU discharge . According to our hospital protocol , all patients presenting to the hospital with tetanus are admitted directly to the ICU . We identified patients with non-neonatal tetanus using either the ICU admission registry or the separate inpatient medical registry . Patients were diagnosed based on clinical findings of rigidity and/or spasms often preceded by a penetrating injury . Medical notes were sought from the medical records department . Detailed information on care was only available for patients after 2008 . The main sources of information were the ICU admission notes , ward round notes and daily charts . We collected data on age , sex , place of residence , time taken to present to hospital , provision of clinical care , time to care provision , and outcomes . We categorized patients into 3 groups based on the year of presentation . The pre-protocol group included patients admitted with tetanus prior to the protocol implementation in 2006 ( 2001–2006 ) . The Early group included patients admitted between 2006 and 2011 , and the Late group included patients admitted between 2012 ( the year in which the hospital tetanus protocol was modified ) and 2016 . The primary study outcomes were care provision for patients , 7-day mortality in the ICU , and overall mortality . The specific care interventions that we examined included administration of immune globulin ( early in the emergency department or late in the ICU ) , surgical wound care , administration of antibiotics , and airway management in the form of mechanical ventilation and tracheostomy placement . Statistical analysis was performed using STATA 14 . 0 ( College Station , Texas , USA ) and all data was anonymized during the analysis . Descriptive analysis of baseline variables was performed to summarize patient characteristics . Categorical variables were described using proportions and continuous variables were described using medians and interquartile ranges . We compared overall mortality between the pre- and post-protocol groups . We explored the differences between the Early and Late groups by chi-squared test for categorical variables and Wilcoxon rank-sum test for continuous variables . Overall and 7-day case-fatality rates were calculated and compared between the groups . We assessed potential factors associated with mortality by univariable and multivariable logistic regression for the Early and Late groups . All statistical tests were performed at a 5% significance level .
A total of 277 patients were admitted to the BMC ICU with tetanus between May 2001 and September 2016 . No cases of neonatal tetanus were admitted; all patients were 8 years older and above . Thirty-one ( 11 . 2% ) patients were admitted before January 2006 and were classified in the pre-protocol group , 133 ( 48 . 0% ) patients were classified in the Early group , and 113 ( 40 . 8% ) were classified in the Late group . Medical files were available for 162 patients ( 59 in the Early group and 103 in the Late group ) . When comparing the pre and post-protocol groups , both were mostly male ( 12/14 ( 85 . 7% ) and 211/246 ( 85 . 8% ) respectively , p = 0 . 995 ) , and there was no significant difference in age ( 30 . 0 [21–42] years and 30 . 0 [19–46] years respectively , p = 0 . 86 , Table 1 ) . Overall case fatality was similar in both the pre- and post-protocol group ( 17/28 ( 60 . 7% ) and 119/238 ( 50 . 0% ) , p = 0 . 28 ) . The case-fatality rate by year of admission is presented in Fig 1 . When comparing the Early and Late groups , for whom detailed medical record data was available , we found no differences between the two groups with respect to sex , age , or distance travelled to reach BMC ( Table 1 ) . The groups had similar median time intervals from the onset of symptoms to hospital presentation , though significantly more people in the Early group presented later ( 3 [1–7] versus 3 [1–4] days , p = 0 . 042 ) . The Early group also experienced significantly more wounds to the back than the Late group . No other significant differences in clinical presentation were observed . Care interventions significantly differed between the two post-protocol groups ( Table 2 ) . We observed significant increases in surgical wound care , initiation of mechanical ventilation and performance of tracheostomies in the Late group compared to the Early group . Both groups had similar in-hospital rates of immune globulin administration , but 57/95 ( 60 . 0% ) of patients in the Late group received early administration of immune globulin in the emergency department compared to 9/50 ( 18 . 0% ) of patients in the Early group ( p<0 . 001 ) . We additionally observed a reduction in the time taken to initiating mechanical ventilation ( 2 . 0 [2 . 0–4 . 0] days in the Early group versus 0 . 0 [0 . 0–2 . 0] in the Late group , p = 0 . 0030 ) and in performing tracheotomies in those for whom mechanical ventilation was initiated ( 12 . 0 [8 . 5–18 . 5] days in the Early group versus 1 . 0 [0 . 0–8 . 0] in the Late group , p = 0 . 023 ) . All patients in the Early and Late group received antibiotics . Contrary to our hypothesis , we did not find a significant reduction in the overall hospital case-fatality ratios between the Early and the Late groups . In contrast , there was a trend towards an increase in mortality from 56/126 ( 44 . 4% ) in the Early group to 63/112 ( 56 . 3% ) in the Late group ( p = 0 . 069 ) . Additionally , there was no change in 7-day case-fatality ratio ( 43/126 ( 34 . 1% ) versus 35/112 ( 31 . 3% ) , p = 0 . 64 ) . The Early group had 6/20 ( 30 . 0% ) deaths attributed to loss of airway , mostly due to laryngospasm , whereas in the Late group , this accounted for only 1/56 ( 1 . 8% ) death ( p = 0 . 001 ) . Sepsis accounted for 3/20 ( 15 . 0% ) of all deaths in the Early group and 25/56 ( 44 . 6% ) in the Late group ( p = 0 . 029 ) ( Table 3 ) . Of note , there were no other clear changes between the Early and Late groups such as ICU staff: patient ratios , nurse training , ICU admission numbers , timeliness of antibiotic administration , or decontamination procedures . All variables listed in Table 1 were analysed as possible predictors of overall ICU mortality in the Early and Late groups and are presented in Table 4 . In the Early group , patients with longer time from symptom onset to presentation at the hospital had lower odds of death compared to patients with shorter time to presentation ( OR = 0 . 68 [0 . 52–0 . 90] for each day delay in presenting for care , p = 0 . 007 ) . Patients for whom mechanical ventilation was initiated had higher odds of death compared to patients not receiving mechanical ventilation ( OR = 4 . 53 [1 . 24–16 . 58] , p = 0 . 022 ) . The increased odds of death may in part have been related to development of sepsis , though numbers were too small to draw conclusions . In the Late group , 25 out of 72 ( 34 . 7% ) patients who were mechanically ventilated developed sepsis , compared with 3 out of 13 ( 23 . 1% ) in the Early group ( p = 0 . 53 ) . In the Late group , older patients as well as patients with longer time from presentation to initiation of mechanical ventilation had higher odds of death ( OR = 1 . 05 [1 . 02–1 . 07] for each increasing year of age , p<0 . 001 and OR = 1 . 63 [1 . 06–2 . 52] for each day delay in receiving mechanical ventilation , p = 0 . 028 respectively ) . Multivariable analysis showed that the factors associated with mortality in the Early group were age ( OR = 1 . 04 [95% CI 1 . 00–1 . 08] , P = 0 . 03 ) , mechanical ventilation ( OR = 26 . 7 [95% CI 2 . 06–274 . 07] , p = 0 . 006 ) and time from symptom onset to attending care ( OR = 0 . 50 [95% CI 0 . 32–0 . 81 ) , p = 0 . 006 ) . In the Late group , the factors that remained significantly associated with mortality were age ( OR = 1 . 03[95% CI 1 . 00–1 . 06] , p = 0 . 01 ) and time from presentation to initiation of mechanical ventilation ( OR = 1 . 74[95% CI 1 . 07–2 . 81] , p = . 024 ) .
Our work demonstrates that the use of a standardized protocol was associated with a significant improvement in the implementation of specific interventions that have been shown to reduce mortality from tetanus . However , this led to an increase in invasive procedures , and therefore the anticipated reduction in overall mortality over a 10-year period was not seen . While the deaths due to respiratory failure and airway obstruction decreased , those due to sepsis increased . This ultimately led to longer ICU stays over the 10-year period without improving mortality . To our knowledge , this is the first study from sub-Saharan Africa to look at the impact of protocolised ICU care for tetanus patients . Our findings suggest the urgent need for additional work to optimize ICU care to help offset the mortality from secondary causes which arise due to gains from improvement in immediate survival . Furthermore , data from our study shows mortality from tetanus remains high and that efforts to improve adult immunization to prevent this neglected disease should be prioritized . Most previous studies have highlighted poor provision of clinical care as a key contributory factor to observed high mortalities . In the Late period of our protocol implementation , patients experienced higher levels of clinical care as compared to other studies . Rates of mechanical ventilation ( 69 . 9% in our study vs 10 . 5% in other studies ) [16 , 17] , administration of immune globulin ( 93 . 1% vs 9–65% ) [17 , 18] , surgical wound care ( 41 . 0% vs 11 . 8% ) [18] , administration of antibiotics ( 100% vs 58% ) [17] , and tracheotomies ( 39 . 2% vs 11% ) [16] were all higher in the Late period of our study compared to rates reported in other studies . A structured approach via a standardized protocol was likely a key contributory factor to the increase in the clinical care . In spite of the protocol , the mortality rates we have reported in both time periods of our study are very similar to rates reported in other studies from similar settings . There is mixed evidence for the effectiveness of protocol standardization for medical care both globally [19 , 20] and also in low- and middle-income countries [21 , 22] . Nonetheless , protocolised care is still common , especially in environments where specialised care is not readily available and where providers’ levels of training and expertise may be variable . Despite highly significant increases in clinical care provision with earlier administration of immune globulin , an increase in surgical wound care , and an increase in mechanical ventilation and tracheostomies , our results showed no improvement in mortality . In fact , we even observed a non-significant trend towards increased mortality in the Late years of implementation of the protocol compared to the Early years ( 56 . 3% versus 44 . 4% , p = 0 . 069 ) . The similarity in demographics between the two groups reinforces the hypothesis that the differences in mortality may be due to the protocol itself . Implementation of the hospital protocol for tetanus management appears to have led to a shift in the causes of death among tetanus patients . In the Early years after protocol implementation , airway obstruction and respiratory failure were the most frequent causes of death , consistent with findings from other studies [3 , 16] . In contrast , sepsis was the most likely cause of death in the Late group , increasing from 15% at baseline to 45% in the Late group . It is likely that increased rates of mechanical ventilation and performance of tracheostomies contributed to a decline in the number of airway/respiratory deaths but that this overall increase in interventions led to an increase in the number of deaths by sepsis , for at least three reasons . First , the Late group underwent more invasive procedures including mechanical ventilation , surgical tracheostomies and surgical wound treatment , each of which increases the risk of hospital-acquired infections ( HAIs ) . Second , our clinical experience suggests that patients in the Late group also more frequently had central venous catheters inserted , which poses an additional risk factor for HAIs . Third , patients in the Late group spent more days in the ICU , often immobilized and at increased risk for other causes of in-hospital mortality such as pulmonary embolism , urosepsis and bedsores . Of note , at present there is no specific aspect of the protocol that focuses on reducing HAIs or other critical illness-associated morbidity . There are approximately 27 studies involving more than 25 , 000 patients with tetanus over a 60-year period issued from the African continent [3] . Most of the reported studies have been done at tertiary level health facilities , similar to our hospital . The median age , male predominance , median time to presentation , clinical presentation , and hospital lengths of stay are similar between our two groups and also similar to other studies of tetanus patients in Africa [3] . Most of the studies reported similar high overall hospital fatalities [3] . All of this suggests that our hospital setting is similar to many others in sub-Saharan Africa and that our finding that tetanus protocol implementation did not improve mortality is likely to be generalizable as well . Our results are to be interpreted in light of some limitations . First , due to our inability to locate medical records for some of the tetanus patients , we were not able to document the care provided in all cases but only to document basic demographic characteristics and outcomes . Furthermore , additional data that would have been informative , such as trends in antibiotic resistance over time and the specific antibiotics administered , were not available . These limitations highlight the complexity of implementing a hospital protocol and the urgent need for additional studies in this area . In summary , we have demonstrated a significant increase in clinical care in accordance with a standardized protocol for the treatment of tetanus patients . The protocol has not led to the anticipated reduction in patient mortality . The unchanged mortality rate , with a shift in causes of death , highlights several key points for consideration when protocols are implemented in resource-limited settings . First , implementation of protocolised care in resource-limited settings is highly complex and requires in-depth monitoring and assessment of patients , staff , and procedures . In our hospital , we are now working to implement infection control policies and determine antibiotic resistance patterns in an effort to decrease HAIs . We will continue to monitor the effect of this intervention and to consider other possible interventions to decrease the mortality of tetanus patients . In addition , management and early recognition of sepsis is extremely complex in resource-limited settings and more surveillance is needed . Finally , we strongly call for an increase in vaccination coverage for at-risk men in sub-Saharan Africa , beginning with the highest-risk groups such as farmers and motorcyclists [5 , 23] , with the aim of eliminating this preventable , lethal disease .
|
Tetanus is a disease characterized by violent , repetitive muscular spasms that are frequently lethal . It rarely occurs in high-income countries due to a highly effective vaccine coupled with a strong immunisation system . Tanzania is a large low-income country in East Africa . Its size and rapidly growing population are straining factors for the immunisation programme . Women typically receive tetanus booster vaccinations during antenatal care , while men remain at risk for tetanus as there is no additional system in place to ensure post-infancy vaccination . In 2006 , our hospital in northwest Tanzania implemented a standardized management protocol for the frequent cases of tetanus that we admit . We assessed the impact of this protocol on the care and survival of these patients , most of whom are between the ages of 20 and 50 years . Although we observed improving clinical care over a 16-year period , this increased care did not result in the reduction in mortality that had been expected . Our work reflects the complexity of this neglected disease , the challenges associated with protocolised care provision in such settings , and the importance of monitoring the effects of interventions following their implementation in low-income countries . Given the devastating personal and economic consequences of tetanus for patients and their families , we also highlight the urgent need to ensure immunisation for these vulnerable men who are the economic backbone for households and the country .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"death",
"rates",
"medicine",
"and",
"health",
"sciences",
"tracheostomy",
"pathology",
"and",
"laboratory",
"medicine",
"immunology",
"surgical",
"and",
"invasive",
"medical",
"procedures",
"health",
"care",
"preventive",
"medicine",
"bacterial",
"diseases",
"sepsis",
"signs",
"and",
"symptoms",
"population",
"biology",
"vaccination",
"and",
"immunization",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"tetanus",
"intensive",
"care",
"units",
"respiratory",
"system",
"procedures",
"proteins",
"hospitals",
"globulins",
"population",
"metrics",
"biochemistry",
"diagnostic",
"medicine",
"health",
"care",
"facilities",
"biology",
"and",
"life",
"sciences"
] |
2018
|
Pre-post effects of a tetanus care protocol implementation in a sub-Saharan African intensive care unit
|
Breast cancer is the most common malignancy in women worldwide . With the increasing awareness of heterogeneity in breast cancers , better prediction of breast cancer prognosis is much needed for more personalized treatment and disease management . Towards this goal , we have developed a novel computational model for breast cancer prognosis by combining the Pathway Deregulation Score ( PDS ) based pathifier algorithm , Cox regression and L1-LASSO penalization method . We trained the model on a set of 236 patients with gene expression data and clinical information , and validated the performance on three diversified testing data sets of 606 patients . To evaluate the performance of the model , we conducted survival analysis of the dichotomized groups , and compared the areas under the curve based on the binary classification . The resulting prognosis genomic model is composed of fifteen pathways ( e . g . P53 pathway ) that had previously reported cancer relevance , and it successfully differentiated relapse in the training set ( log rank p-value = 6 . 25e-12 ) and three testing data sets ( log rank p-value<0 . 0005 ) . Moreover , the pathway-based genomic models consistently performed better than gene-based models on all four data sets . We also find strong evidence that combining genomic information with clinical information improved the p-values of prognosis prediction by at least three orders of magnitude in comparison to using either genomic or clinical information alone . In summary , we propose a novel prognosis model that harnesses the pathway-based dysregulation as well as valuable clinical information . The selected pathways in our prognosis model are promising targets for therapeutic intervention .
Breast cancer is the second ( after skin cancer ) most frequently diagnosed cancer in women , and ranks second ( after lung cancer ) in the deaths of women in year 2013 [1] . Most clinical studies categorize breast cancer into four molecular subtypes: Luminal A , Luminal B , Triple Negative/Basal like and Her2 [2] , [3] . The survival outcomes differ significantly among the clinical subtypes . Luminal A and B subtypes have a relatively good prognosis , whereas triple negative or basal like tumors , and Her2 tumors have very poor prognosis with much higher recurrence and metastasis rates [2]–[4] . Furthermore , it is increasingly being realized that breast cancers are much more heterogeneous diseases than what is determined by the clinical subtypes , and that better prediction of prognosis is needed early on for more personalized treatment and management . Towards this goal , prognosis biomarkers of breast cancers have been investigated in many studies [5]–[7] , based on signatures from high-throughput platforms such as gene expression profiles . Some signature panels such as the NKI 70 test are currently in commercial use with decent prediction of metastasis [8] . However , transcriptomic data are usually poorly dimensioned with many more genes than the number of samples , thus methods that reduce the dimension by incorporating higher-order information of functional units , such as gene sets , pathways and network modules , have been recently explored [9]–[16] . This methodology is based on the observation that multiple genes involved in the same biological processes are often dysfunctional all together in cancers [17] , therefore features selected from representative functional units are presumably more robust with better biological annotations [10] , [17] . Currently , two main approaches to define functional units have been proposed . One approach is to identify de novo functional units from the data . For example , van Vliet used an unsupervised module discovery method to identify gene modules , scored them and use them as features in a Bayes classifier [18] . Teschendorff et al . reported improved prognostic classification of breast cancers via a novel strategy to discover the activated pathways from the modules of “expression relevance network” [12] . Similarly , network analysis with combination of all the useful gene information has been developed and utilized to measure the coordination among the genes [13] . The other main approach uses the existing pathway information to build functional units . For example , Lee et al used the MsigDB C2 gene sets to select feature sets using the t-test , and represented the pathway activity level by a subset of genes whose combined expression delivered optimal discriminative power for the disease phenotype [14] . Abraham et . al used a set statistic that aggregated the expression levels of all genes in a set , and constructed prognostic gene sets that were as predictive as individual genes , yet more stable and interpretable within the biological context [9] . However , most of these methods model the prognosis as binary outcomes , and post hoc analyze the performance of the methods using survival information; or individualized information of pathway deregulation is lost during information extraction before deriving statistical metrics . More importantly , the merits of combining clinical features and genomic features together have not been adequately addressed in most studies , where the models were only built upon the genomic information . In this study , we use a novel pathway-based deregulation scoring matrix to transform the gene-based genomic features in combination with the Cox regression and L1-LASSO regularization to model survivals . With this pathway deregulation score matrix as inputs , we constructed a pathway-based genomic model consisting of fifteen cancer relevant pathways that successfully predicted relapse difference ( log rank p-value = 6 . 25e-12 , and AUC = 0 . 80 ) and validated them on three breast cancer data sets with diversified clinical profiles ( log rank p-value<0 . 0005 , and average AUC = 0 . 68 ) . The pathway-based genomic models consistently performed better than gene-based models on all four data sets . Moreover , combining genomic level information with clinical information improved prognosis prediction and classification by at least three orders of magnitudes of p-values , in comparison to either genomic or clinical information alone .
We used four individual gene expression microarray data sets for the testing and validation of the pathway-based prognosis model ( Table 1 ) , all of which were measured by Affymetrix HG-U133A array and had relapse and survival information . We used the data set of 236 patients in Miller et . al . [19] as the training data mainly because this data set contains the most abundant clinical information , including ER status , PG status , tumor size , grade , lymph node status and P53 mutation . PAM50 is a list of 50 genes initially proposed to successfully differentiate the breast cancer subtypes and it was later found that PAM50 also harbors good prognosis information on breast cancer [20] . Therefore , we first present the testing data summary results and correlate relapse with PAM50 and other clinical factors ( Figure 1 ) . Although tumor molecular subtypes are unknown due to the missing Her2 marker information , we nevertheless observed a good correlation between PAM50 matrix and relapse . Based on the hierarchical clustering results of PAM50 heatmap , we dichotomized the samples into high and low risk groups , This grouping approach , without any supervised learning , results in a fairly good association to relapse status ( Chi-square test p = 7 . 46e-5 ) . Additionally , grade and lymph node have significant associations to relapse , with Chi-square test p-values of 0 . 018 and 9 . 146e-6 respectively . Single clinical factor based survival analysis also confirms such significant relevance to relapse: p-values of Wilcoxon log rank tests for the p53 , grade , tumor size and lymph node status based survival differences are 0 . 0152 , 0 . 00181 , 1 . 92e-7 and 4 . 93e-8 , respectively . Similar to previous observations [21] , ER and PG status are not good prognosis indicators , with the log rank test p-values of 0 . 819 and 0 . 227 , respectively . There are a total of around 600 samples in the three testing data sets , 2 . 5 times the size of samples in the training set . Testing set 1 ( Ivshina data ) [22] and testing set 2 ( Pawitan data ) [23] have very similar distribution pattern to the training data ( Miller data ) [19] . However testing set 3 ( Desmedt data ) [24] has very different distribution compared to other three data sets , as the samples were all lymph node negative tumors . We include set 3 as an extension to the other two testing data sets to exam the performance of the pathway-based genomic model for prognosis . We have developed a novel pathway-based prognosis prediction model , unlike most other models that are gene-based ( Figure 2 ) . We transformed a conventional gene-based matrix into a new pathway-based matrix of reduced numbers of rows , where each row represents a KEGG or BIOCARTA pathway-based scores over all samples ( columns ) . Instead of using log2 transformed intensities as elements of the matrix , we used Pathway Dysregulation Scores ( PDS ) [25] that measure the distance of a particular pathway to the “normal condition” curve in a hyperspace . PDS ranges from 0 to 1 , and the higher PDS score signifies more “abnormity” . This pathway-based PDS matrix was used as the initial input to select featuring pathways that are predictive of survival , based on the multi-variate Cox-PH model [26] . We used L1-LASSO penalization method [27]–[29] to constrain the featuring pathways to be selected . To be consistent , we conducted 250 simulations to select the best set of pathways . We first evaluated the featuring pathways selected by the model , in relation to other clinical factors and relapse status in the training data set ( Figure 3 ) . Comparing the heatmap of selected featuring pathways to that of the PAM 50 genes ( Figure 3A ) , the selected pathways are more prognostic for relapse . This is supported by two observations: ( 1 ) Dichotomized samples of high risk and low risk groups through hierarchical clustering of PDS scores have a higher correlation to relapse status ( Chi-square test p = 1 . 99e-6 ) , compared to those of PAM50 gene matrix ( Chi-square test p = 7 . 46e-5 ) and ( 2 ) The median PDS scores over fifteen selected pathways have a correlation coefficient of 0 . 17 to relapse , in comparison to 0 . 08 for the median expression intensities over PAM50 genes . Thus the selected pathways by our model are better prognostic features than PAM50 genes , in terms of the correlation to disease relapse . To investigate the performance of the model , we used the PI value which is the logarithm of hazard ratio from the fitted Cox-PH model to dichotomize the samples , similar to others [21] [30] . We divided the samples into higher and lower risk groups with a 3 to 1 ratio ( 3rd quartile in PI ) , in order to match the relapse versus non-relapse sample ratio in the training data . Samples with larger PDS scores are expected to have higher PI scores , and are more likely to have relapsed diseases . The same PI threshold was applied to dichotomize the training data set as well as multiple independent testing data sets . The performance of the genomic model was then evaluated by two approaches: ( 1 ) the Wilcoxon log rank test p-values of the Kaplan-Meier survival curves from the two risk groups in each data set , and ( 2 ) the AUCs of ROC curve based on binary classification . Instead of combining all four data sets for meta-analysis , we kept them as individual data sets to validate the robustness of our model . As expected , the pathway-based genomic model is highly accurate at differentiating the risks of breast cancer relapse within the training data , with a Wilcoxon log rank p-value of 6 . 25e-12 ( Figure 4A ) . The model yields very decent predictive results with the p-value of 1 . 52e-4 in testing set 1 and 3 . 91e-5 in testing set 2 ( Figure 4B and 4C ) . The predictive performances are expected to drop in the testing data sets , since they have different patient populations and clinical characteristics from the training set ( Table 1 ) . Impressively , the model gives a very significant p-value of 3 . 73e-4 for testing data set 3 ( Figure 4D ) , which are all early stage lymph node negative tumors whose prognosis is very difficult to predict . Additionally , we evaluated the performance of models using binary classification . We used the relapse/non-relapse information in the data sets as truth measures , and the model's high vs . low risk classification as predictions . As shown in Figure 4E , the ROC curve in the training set gives an AUC value of 0 . 80 , and AUCs of 0 . 73 ( testing set 1 , Pawitan data ) , 0 . 67 ( testing set 2 , Ivshina data ) , 0 . 65 ( testing set 3 , Desmedt data ) , consistent with the results in Kaplan-Meier curves ( Figure 4A–D ) . To examine the effect of total number of input pathways on model performance , we randomly kept 1/2 , 1/4 , 1/8 and 1/16 of all input KEGG and BioCarta pathways in the training dataset , and then generated the PDS Matrices for 18 simulations under each scenario . For each simulation , we built the model with the same workflow as in Figure 2 and computed the Wilcoxon log-rank test p-value between the survival curves of the two risk groups , as well as the AUCs of the classification results . The boxplot in Figure S1 shows a gradual decrease of AUCs due to the input pathways , in the order of 1/2>1/4>1/8>1/16 pathway-based models . The difference between 1/2 and 1/4 pathways is significant ( p-value<0 . 05 ) . All AUCs , however , are in the range between 0 . 69 and 0 . 81 . Our earlier results of selected pathway features vs . PAM 50 genes suggested that pathway-based features may be better than gene-based features . To validate this , we trained the four data sets individually and compared within the same data set the performance of pathway-based models and gene-based genomic models which do not have the PDS matrix generation step ( Figure 2 ) . In order to test the risk differentiation power of the model , the cutoff PI value in each data set was set to match the ratio of relapse vs . non-relapse patients in that particular set . The results of Kaplan-Meier survival curves and ROC plots based on classification all consistently show that pathway-based genomic models are superior to the gene-based models ( Figure 5A–H ) . For example , in Miller data set the log-rank p-value is 6 . 25e-12 for the pathway-based model ( Figure 5B ) , compared to that of 1 . 75e-9 for the gene-based model ( Figure 5A ) . In the Desmedt data set , the p-value of the pathway-based model is even more significant than that of gene-based model ( 5 . 12e-36 vs . 8 . 84e-12 , Figure 5H and 5G ) . Similarly , pathway-based genomic models have better ROC curves than gene-based genomic models ( Figure 5I ) , with AUCs of 0 . 80 vs . 0 . 78 in Miller data , 0 . 85 vs . 0 . 77 in Pawitan data , 0 . 74 vs . 0 . 70 in Ivshina data , and 0 . 92 vs . 0 . 76 in Desmedt data . To estimate the statistical significance of comparisons among the pathway-based and gene-based models , we performed leave-one-out cross validation ( LOOCV ) simulations to compute the Wilcoxon log-rank test p-values and AUCs of ROC classification curves . The cross validation results show that statistically the pathway-based models perform better than the gene-based models ( Figure S2 , all t-test p-values<0 . 001 ) . These results are consistent with the observations from previous studies [12] , [14] , and support the hypothesis that including higher-order secondary information yields better prognostic values . NKI70 ( Mammaprint ) is one of the most commonly used model for breast cancer prognosis prediction , and it has been approved by FDA for commercially use in clinics . To demonstrate the potential clinical utilities of our model , we compared the NKI70 method with ours , and applied the NKI70 method to our training data set ( Miller data ) . We first mapped the NKI70 gene signatures [8] to the genes in the U133A array , then correlated the gene-expression profile with the good-prognosis/poor prognosis data from the NKI study and classified the samples into good and poor clusters as done previously [7] . The NKI70 test gives a Wilcoxon log-rank test p-value of 2 . 58e-3 for the survival analysis , in contrast to the p-value of 6 . 25e-12 obtained by our pathway-based model; it only yields an AUC of 0 . 62 for classification , in contrast to 0 . 80 from our model ( Figure S3 ) . Previous studies suggested that clinical information of breast cancers provides additional values to a genomic model that was built on lists of genes [21] . To test if such merit of clinical information also applies to our genomic model of fifteen pathway features , we investigated the performances of the genomic , clinical and genomic-clinical combined models . Since the scales of PDS and clinical features vary significantly , we re- normalized PDS and clinical features independently to have the standard normal distribution , so that they are subject to the same selection criteria . The resulting clinical model is composed of four selected features: grade , tumor size , p53 and lymph node . This is not surprising , as they are also significant factors in the univariate Cox-PH models ( Table 2 and Figure 1B–E ) . The combined model keeps ten of the fifteen pathways ( Table 2 ) and about 60% of genes that were selected by the genomic model . It also selects tumor size and lymph node status as additional features ( Table 2 ) . This is expected given their highly significant p-values ( 1 . 92e-7 and 4 . 93e-8 , respectively ) in the univariate Cox-PH models ( Figure 1B and 1E ) , as well as relatively large coefficients in the clinical model ( 0 . 27 and 0 . 36 , respectively ) . Since only testing data set 2 has both tumor size and lymph node information , we used this data set and the testing data set to demonstrate the performances of genomic , clinical , and combined models . The comparisons present the compelling advantage of combining clinical and genomic information in a model ( Figure 6 ) . As shown in the training data , selected clinical features are undoubtedly important: the Wilcoxon log rank test p-value of the clinical model is 2 . 21e-10 ( Figure 6E ) , slightly less significant than the pathway-based genomic features by two orders of magnitude . Most importantly , the combined model is much better than either genomic model ( p-value = 6 . 25e-12 ) or clinical model alone , with a p-value of 1 . 88e-24 ( Figure 6C ) . This trend of significances is consistent in the testing set 2 , with the p-values of 1 . 12e-7 in the combined model ( Figure 6D ) , 1 . 52e-4 in the genomic model ( Figure 6B ) , and 2 . 7e-3 in the clinical model ( Figure 6F ) . Moreover , the ROC curve comparisons of these three models also show the same order of performances: combined model>genomic model>clinical model , with AUCs of 0 . 83 , 0 . 80 , and 0 . 74 in the training set , and 0 . 71 , 0 . 68 and 0 . 65 in the testing set 2 ( Figure 6G ) . To demonstrate the statistical significance of comparisons among the pathway-based , clinical and combined model in the training set and the testing set 2 , we performed leave-one-out cross validation ( LOOCV ) simulations to compute the Wilcoxon log-rank test p-values and AUCs of ROC classification curves . The cross validation results show that statistically the combined model performs better than the pathway-based model , and the pathway-based model performs better than the clinic model ( Figure S4 , all p-values<0 . 001 between pathway-base/clinical models and combined models ) . We expect that the consensus pathways selected both in our genomic model and combined model convey important cancer-related functions . To test this we examed the annotations of this subset of ten pathways ( Table 2 ) . Interestingly , KEGG_MELANOGENESIS is selected as a feature , probably due to inclusion of many cancer relevant genes in this pathway: such as protein kinase genes PRKACB , PRKACG , PRKCB , PRKCA; phosphorylase kinase genes CALM1 , CALM2 , CALM3; G-protein related gene GNAQ , HRAS; mitogen-activated protein kinases MAPK1 , MAPK3 , MAP2K1; and other oncogenes like RAS [31] , [32] . Many of these genes have been shown to function in breast cancer progression [31] . Impressively , multiple signaling pathways are selected , including BIOCARTA_P53_PATHWAY , BIOCARTA_SRCRPTP_PATHWAY , BIOCARTA_PYK2_PATHWAY , BIOCARTA_VIP_PATHWAY , BIOCARTA_RARRXR_PATHWAY , and BIOCARTA_AKAP13_PATHWAY . They are well-known to be associated with breast cancers prognosis [33]–[39] . The best example is BIOCARTA_P53_PATHWAY , the dysregulation of p53 Signaling Pathway is well-documented , and the tumor-suppressor gene p53 has one of the highest mutation rates in breast cancer [5] , [19] . In addition , some pathways related to basic cell functions are selected as prognostic features . For example , G1_PATHWAY is selected , and the G1/S cell cycle checkpoint controls are well known to be dysfunctional in many cancers including breast cancer [40] . FATTY_ACID_METABOLISM is also selected by the model , and many studies have showed that fatty acid metabolism is involved in breast cancer [41] . In particular , Fatty acid synthase ( FASN ) is highly expressed in breast cancer with a poor prognosis compared to others [41] . Interestingly , BIOCARTA_RNA_PATHWAY is also selected , largely due to its members TP53 and MAP3K14 that are closely related to breast cancer . A total of 265 genes are overlapped between the selected pathways of the genomic model and the combined model . Table 3 summarizes the top 30 genes that are involved in the selected pathways . They are ranked by weighted sum of both occurrences in selected pathways ( counts ) and weights measured by the hazard ratio of each pathway . Among them , many genes encode protein kinases that are well-known to be involved in breast cancers , such as PRKACB , PRKACG , MAPK1 and CALM1 . Some other genes encode transcription factors that are well-known for their close relationship to cancer , such TP53 , RB1 , HRAS , RAF1 , GRB2 , E2F1 , and SRC [32] , [42]–[44] . We therefore conclude that the selected pathways are prognostic features of significant cancer relevance .
The heterogeneity of cancers is being increasingly recognized , suggesting more personalized care decisions with treatment for individual patients are needed . As a result , prognosis prediction of breast cancers with high-throughput data has been a growing topic in recent years . Many statistical and machine learning methods have been developed to analyze various types of high-throughput cancer genomics data , by taking advantage of higher-order relationships among genes . The hypothesis is that the highly correlated gene-based markers often represent identical biological processes; therefore by including higher-order representative features , such as Gene Ontology sets , pathways and network modules , the prediction will be more stable [9]–[14] , [45] . Our novel method of prognosis prediction presented in this study belongs to this class of methods . However , unlike some other methods where individual pathway information is lost due to summarization or transformation , the pathway features proposed in this study explicitly measure the degrees of pathway dysregulation for cancer recurrence . Comparing selected pathways and the PAM50 genes which were demonstrated to be prognostic [20] , the PDS-based pathway approach has better correlation to breast cancer relapses . Moreover , when comparing gene-based with the pathway-based genomic models , where the only difference between them was the input matrix , pathway-based models uniformly performed better than gene-based models in all the data sets we tested . Our results are consistent with several other gene-set/pathway-based models [9] , [14] , where different summarization metrics were used . It will be very interesting to compare the prediction results based on these different metrics in a follow-up study . To demonstrate the robustness in predicting differential risks of relapse from the pathway-based genomic model , we chose to train and test on independent study samples , rather than combining them together as a large data set [21] , [46] , which would diminish the effect of population heterogeneity . Despite population difference and much bigger testing data size relative to the training data size , the method still achieved good performance on all three testing data sets , including a data set of all early stage lymph node negative tumors where prognosis is particularly difficult to predict . Another merit of our method is that it enables combining the important clinical information with the pathway-based genomic information . Even though the clinical model by itself is the least predictive , compared to the genomic model and the combined model , it is nevertheless significant and informative , as shown by tumor size and lymph node status . The genomic model is better than clinical model alone . However , the combined model of clinical and genomic features performs the best . Our conclusions agree and extend the earlier work from Fan et al . [21] who focused on prognosis prediction of all node-negative and systemically untreated breast cancer patients , since we include both node-negative and node-positive samples . The results of the genomic model ( AUC = 0 . 80 and p-value = 6 . 25e-12 in training data , and AUC = 0 . 68 and p-value = 1 . 52e-4 in test data 2 ) and the combined model ( AUC = 0 . 83 and p-value = 1 . 88e-24 in the training set , and AUC = 0 . 79 and p-value = 1 . 12e-7 in test data set 2 ) are better than what was recently reported by Vilinia S et al [47] . They obtained an AUC = 0 . 74 for the training set and 0 . 65 for the testing set , in a model that combined signatures of mRNA and microRNAs deriving from the TCGA IDC cohort sequencing data . This suggests the advantages of combining PDS based pathway score inputs with a Cox-PH model and LASSO penalization approach: even though the genomic data in our study are based on microarrays that have more noise and smaller sample sizes , they still yield better predictive results in comparison to the combined mRNA and microRNA sequencing signatures obtained from a larger sample size . It will be of great interest to apply our models to the TCGA breast cancer mRNA and microRNA sequencing data in the future . The pathways selected by the model show biological relevance to breast cancer prognosis . The fatty acid metabolism pathway is found to be crucial to maintain the cancer cell malignant phenotype , and higher expression of fatty acid synthase has been discovered as a common phenotype in breast cancer with a poorer prognosis [41]; As another example , Src kinase activation by protein tyrosine phosphatase alpha ( SRCRPTP_PATHWAY ) , has been discovered in invasive breast cancer with compelling evidences . Src inhibitors are being considered as potential therapy to treat invasive breast cancers , as inhibition of c-src was recently found to be involved in E2-induced stress which would finally result in apoptosis in breast cancer cells [33] . Increasing evidence shows that vasoactive intestinal peptide ( VIP ) in BIOCARTA_VIP_PATHWAY is highly expressed in breast cancer cells along with its receptor [33] , and VIP-targeted nanomedicine is under study as therapy for breast cancer [34] . Pyk2 in BIOCARTA_PYK2_PATHWAY is linked to map kinases MAPK , which has wealthy records in breast cancer studies [35] . RARRXR_PATHWAY is the RAR/RAR nuclear receptor complex that is co-activators to facilitate initiation of transcription in carcinoma cells [37] . And BRX , the truncated form of Rho-Selective Guanine Exchange Factor AKAP13 in the BIOCARTA_AKAP13_PATHWAY , has been identified to function as an ER cofactor [39] . Although the workflow proposed in this study is generic and the pathway features are clearly significant , we should point out a few potential limitations of the model . First of all , the pathway-based model is trained and tested on gene expression data from the U133A platform . We suspect that direct application of the model to other platforms , such as RNA-Seq , is not desirable , and some additional re-processing work has to be done additionally . The reason is that data distributions maybe very different between various platforms . One notorious example is that biomarkers identified by high-throughput microarray platform often had poor correlations in qPCR platform . Thus we recommend that when researchers use the workflow in Figure 2 on different data types , they may increase the predictive power by retraining the model with their own data . Another limit of our approach is that we only used the information from genes that compose the 403 pathways that we considered , thus some gene-level information is unavoidably lost . In our case , over 4500 genes were enlisted in the pathways , and among them over 3200 genes are probably expressed ( averaged log 2 expression intensities >7 ) . On the other hand , the raw U133A array has results of over 14 , 000 genes within which over 10 , 000 genes are probably expressed . Therefore our model captures about 1/3 of the gene-level information overall . One can certainly use other curated gene sets , such as the MsigDB C2 gene sets , to increase the coverage of the genes by the pathways . However , from the sensitivity analysis that we have performed ( Figure S1 ) , we only observed a slight decrease of model performance based on AUCs , which are in the range of 0 . 69 and 0 . 81 . In conclusion , we propose a novel pathway-based genomic model that measures the pathway-based deregulation score and shows significant prognosis values . This pathway-based genomic model performs better than the gene-based genomic model . Additionally , we found that combining the clinical information of lymph node status and tumor size improves the performance of the prognosis model . Many selected pathways in our study present values for breast cancer prognosis prediction , and they are also promising therapeutic targets for future investigations .
We used four publicly available data sets of breast cancer samples from National Center for Biotechnology Information ( NCBI ) Gene Expression Omnibus ( GEO ) GSE4922 [22] , GSE1456 [23] , GSE3494 [19] and GSE7390 [24] . All four data sets are based on Affymetrix HG-U133A microarray platform , and have relapse-free survival information as well as some other clinical information , as shown in Table 1 . For data set GSE7390 [24] , all patients are lymph node negative . The GSE3494 data set was used as the training set as it has more clinical information , and all others were used as testing data sets . We mapped original probe IDs to Gene IDs using R package biomaRt [48] . In order to relate the probe ID to the Gene ID , we downloaded the array annotation file and used the RefSeq IDs as the intermediates to map to the Gene ID . When a gene has multiple probes , we computed the geometric mean of log2 transformed probe intensities as the gene expression . All the data sets were normalized independently between array using limma package [49] . To minimize batch effects across different data sets , we used the CONOR package with the Bayesian method [50] . We generated the PAM50 heatmap of the gene expression data and the correlation heatmap with hierarchical clustering , where Euclidean distance measure was employed . For the clinical factors , we correlated their associations with the relapse in the training data set with both Chi-square test and Wilcoxon log-rank test for survival curves . The pathway information was obtained from the GSEA ( http://www . broadinstitute . org/gsea/ ) curated gene sets that include a total of 403 pathways from Biocarta ( http://www . biocarta . com ) [51]and KEGG [52] . To perform gene sets analysis , we used R package Pathifier [25] , an algorithm that transforms the information from the gene level to pathway level and infers pathway deregulation scores for each pathway within each sample . The pathway deregulation score ( PDS ) in each sample is a measure of degrees of the deviation of a specific pathway from the “normal status” located on the principle curve . The concept of principle curve was proposed by Hastie and Stuetzle [53] as a nonparametric nonlinear extension of the PCA ( Principle Component Analysis ) in which the assumptions of dependence in the data are avoided . A principle curve is a one-dimensional curve that is derived from the local average of p-dimensional points and goes through the cluster of p-dimensional principle components . It sensibly captures the information of variation in all the samples . Specifically , the single parameter λ varies tracing the whole data along the curve [53] . The curve f ( λ ) is defined to be a principal curve if for arbitrary λ . The principle curve is built through iterations of smoothed procedure in the local average of data points . If one sample differs from others in one specific pathway , the distance to the curve is further and it leads to a higher PDS score and vice versa . In the model selection stage , we used Cox-Proportional Hazards ( Cox-PH ) model based on L1 – penalized ( LASSO ) estimation [27]–[29] , with the R package penalized [29] . With the input of both PDS score containing the gene sets information and survival information of time and relapse , a tuning parameter lambda was used to restrict the number of parameters in the model . The optimal lambda was selected after running 250 simulations through likelihood cross-validation . A prognostic genomic model was thus generated with specific pathways and coefficients . We then computed a Prognosis Index ( PI ) score which is the logarithm of hazard ratio . We divided the samples into two groups of higher risk and lower risk with a 3 to 1 ratio , based on the 3rd quartile of PI . We used this cutoff to reflect the relapse/non-relapse ratio in the training data set . We tested the above model in three other data sets . To do so we used the same PI cutoff above and separated samples into predicted high risk and low risk groups . We then used Kaplan-Meier curve together with Wilcoxon log rank test to evaluate the performance of our model . To generate the receiver operating characteristic ( ROC ) curves , PIs are used as predicted values in comparison to the “truth” values of relapse/non-relapse information . The confusion matrix with sensitivity and 1- specificity is calculated for each division in ROC curves and the areas under the curve ( AUC ) is shown along with the ROC plot . To determine whether the clinical factors improve the prognosis of genomic pathway-based model , we re-normalize the clinical factors and molecular PDS independently to ensure that each factor has the standard normal distribution . We then combined the normalized clinical and molecular factors into the LASSO penalized step and built the combined model using the optimized lambda through 250 simulations , similar to the construction of the genomic model as described earlier . The model performance comparisons were also done similarly to those of the genomic model . We used survival analysis to compare the relapse-free-survival results in the training and testing data sets . Patients without these events during the study were considered censored . We used the Cox-PH model to associate the risk of relapse to selected pathway features and clinical features by L1- LASSO . The Cox model is a semi-parametric model that is widely used to analyze the survival data . The non-parametric portion comes from the fact that no assumptions are made about the form of the baseline hazard . However , it has the assumption that the log hazard ratios are constant over the time for each feature . Assume that we obtained p features to be related with breast cancer relapse for each patient , Cox-PH model represents the relationship between the risk of relapse and X features as:Here is the baseline hazard ( instantaneous risk ) which only depends on time . The ratio of hazard ( HR ) between two pathway or clinical features and is:The relative hazard between any two features is constant over time and only depends on the differences of the values in features . The PI for each patient J's features is calculated asThis risk factor can be easily transformed to hazard ratio for different features , assuming that we have a baseline feature . The weights for different features were calculated from the training data set using the Cox-LASSO model . For the genomic , clinical and combined models , we used Kaplan Meier curves to present the prognosis performance in classified high risk and low risk groups . The data set was dichotomized into two groups , and the higher risk group is assumed to have higher hazard of relapse compared to the lower risk group . We used the Wilcoxon log-rank test to check the survival difference between these two groups . To find the significance of an individual factor's impact on relapse , we fit individual predictor with a univariate Cox-PH model . We then calculated the hazard ratio by computing the exponential of the coefficients in the Cox-PH model . All survival analysis was conducted using the R package Survival [54] . To examine the effect of input pathways on model performance , We randomly select 1/2 , 1/4 , 1/8 and 1/16 of all input KEGG and BioCarta pathways , then generated the PDS Matrices for 18 times under each case . For each simulation , we built the model with the workflow in Figure 2 and computed the Wilcoxon log-rank test p-value between the survival curves of two risk groups , as well as the AUC of the classification results . We then used boxplots to demonstrate the differences of –log10 ( p-values ) and AUCs due to different total pathway counts . To estimate the statistical confidence of comparisons of each model , we used leave one out cross validation ( LOOCV ) to compute p-values and AUCs across all simulations . In the ith simulation ( i = 1 , … , total sample size of the data set ) , we deleted the ith patient sample , modified the PI threshold by the remaining sample ratio of recurrence to non-recurrence and finally calculated the Wilcoxon log-rank test p-value as well as the AUC of the classification results . We then used boxplots to demonstrate the comparisons between the pathway-based and the gene-based models , and among the genomic , clinical , and the combined models . We tested the NKI70 method to our training data set ( Miller data ) . We mapped the NKI70 gene signatures from to the genes in the U133A array . We correlated the gene-expression profile with the good-prognosis/poor prognosis data from the NKI study [8] , and then classified the samples into good and poor clusters as done by others [7] . For consistency , we used the Wilcoxon log-rank test p-value from survival analysis and the AUC of the ROC classification to assess the results .
|
With the increasing awareness of heterogeneity in breast cancers , better prediction of breast cancer prognosis is much needed early on for more personalized treatment and management . Towards this goal we propose in this study a novel pathway-based prognosis prediction model , which emphasizes on individualized pathway-based risk measurement using the pathway dysregulation score ( PDS ) . In combination with the L1-LASSO penalized feature selection and the COX-Proportional Hazards regression model , we have identified fifteen cancer relevant pathways using the pathway-based genomic model that successfully differentiated the relapse in the training set as well as three diversified test sets . Moreover , given the debate whether higher-order representative features , such as GO sets , pathways and network modules are superior to the gene-level features in the genomic models , we demonstrate that pathway-based genomic models consistently performed better than gene-based models in all four data sets . Last but not least , we show strong evidence that models that combine genomic information with clinical information improves the prognosis prediction significantly , in comparison to models that use either genomic or clinical information alone .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"bioinformatics",
"epidemiology",
"database",
"and",
"informatics",
"methods",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"disease",
"informatics",
"research",
"and",
"analysis",
"methods"
] |
2014
|
A Novel Model to Combine Clinical and Pathway-Based Transcriptomic Information for the Prognosis Prediction of Breast Cancer
|
The master circadian clock in fish has been considered to reside in the pineal gland . This dogma is challenged , however , by the finding that most zebrafish tissues contain molecular clocks that are directly reset by light . To further examine the role of the pineal gland oscillator in the zebrafish circadian system , we generated a transgenic line in which the molecular clock is selectively blocked in the melatonin-producing cells of the pineal gland by a dominant-negative strategy . As a result , clock-controlled rhythms of melatonin production in the adult pineal gland were disrupted . Moreover , transcriptome analysis revealed that the circadian expression pattern of the majority of clock-controlled genes in the adult pineal gland is abolished . Importantly , circadian rhythms of behavior in zebrafish larvae were affected: rhythms of place preference under constant darkness were eliminated , and rhythms of locomotor activity under constant dark and constant dim light conditions were markedly attenuated . On the other hand , global peripheral molecular oscillators , as measured in whole larvae , were unaffected in this model . In conclusion , characterization of this novel transgenic model provides evidence that the molecular clock in the melatonin-producing cells of the pineal gland plays a key role , possibly as part of a multiple pacemaker system , in modulating circadian rhythms of behavior .
Numerous aspects of animal behavior and physiology vary dramatically during the course of the day-night cycle as an adaptation to recurring environmental fluctuations [1] . These variations are driven by an endogenous timing mechanism , the circadian clock , which is adjusted by external signals such as light to ensure its synchronization with the solar day [2] . At the heart of the molecular clock in vertebrates are daily oscillations in the expression and function of evolutionarily conserved clock genes and their protein products , including CLOCK and BMAL , which form heterodimers that activate the transcription of clock and clock-controlled genes ( CCGs ) via E-box enhancers [3] . Despite high conservation of the molecular architecture of the circadian clock , there are marked differences in organization of the circadian system among different classes of vertebrates . In mammals , neurons of the suprachiasmatic nucleus ( SCN ) were defined as the master circadian clock , based on the findings that: a ) locomotor activity rhythms , a traditional indicator of circadian clock function , are completely lost upon SCN lesion , and b ) light-induced phase shift and entrainment of behavioral rhythms to the light-dark ( LD ) cycle depend on photic input to the SCN [4 , 5] . Photic input acts on the SCN through the retinohypothalamic tract to synchronize rhythmic neuronal activity [2 , 4 , 5] . Signals from the SCN regulate circadian rhythms of multiple targets , including synthesis of the hormone melatonin in the pineal gland [6] and the synchronization and coordination of cell-autonomous molecular circadian oscillators , known as peripheral clocks , found in most tissues [7 , 8] . The mammalian pineal gland is regulated by the SCN via a multisynaptic pathway that leads to night-time sympathetic norepinephrine induction of melatonin synthesis [6] . Rhythmic melatonin production constitutes a major element of the circadian system in mammals , as it contributes to the regulation of various daily and annual physiological rhythms [9] . In birds , rhythmic melatonin synthesis in the pineal gland is also regulated by the SCN via sympathetic innervation [10] . However , the avian and mammalian pineal glands differ in that the avian pineal gland can also function independently to produce melatonin rhythms which are driven by a pineal gland-intrinsic clock , and in that light directly induces phase shifts of these rhythms [11–15] . Accordingly , the avian pineal gland continues to produce circadian rhythms of melatonin when kept in culture without photic cues , and light can act directly on the cultured gland to modulate these rhythms . The functional importance of the avian pineal clock is evident from the finding that pinealectomy leads to physiological and behavioral arrhythmicity . Moreover , rhythmicity can be restored by timed melatonin administration or by pineal transplantation that confers the donor's circadian phase on the recipient [16–18] . Accordingly , the circadian system of birds is considered to consist of multiple pacemakers located in the SCN , pineal gland and retina , which interact to regulate downstream physiological and behavioral rhythms [10] . Hence , the hierarchical organization of the mammalian circadian systems is not a feature shared by all vertebrates . The fish pineal gland , as in birds , is photoreceptive and functions independently . It contains an intrinsic circadian oscillator that drives rhythms of melatonin production even when maintained in culture , disconnected from any neuronal input . Furthermore , the fish pineal gland includes cells with retinal cone photoreceptor-like characteristics , and light directly induces a phase shift of the melatonin rhythms [19–25] . The autonomy of the fish pineal gland is further emphasized by the fact that in some fish species norepinephrine does not affect melatonin synthesis [26 , 27] . Thus , the fish pineal gland presents all the features of a complete circadian system , comprising a photoreceptive pathway , a molecular oscillator , and an overt rhythmic output ( melatonin biosynthesis ) . For these reasons , and since a functional counterpart to the mammalian or avian SCN has not been identified in fish , research has focused on the fish pineal gland as a master clock organ . It should be noted , however , that studies employing pinealectomy and exogenous melatonin administration have yielded species-specific effects , pointing to variations in the contribution of the fish pineal gland and melatonin to the coordination of circadian rhythms of physiology and behavior [21 , 24 , 28 , 29] . In zebrafish , the outcomes of pharmacological and genetic manipulations suggest that melatonin is required for the circadian regulation of the sleep/wake cycle in this species [30–33] . Nevertheless , the presence of peripheral oscillators that are photoreceptive and directly entrainable by exposure to light [34 , 35] has led to the view of a decentralized zebrafish circadian system , thereby questioning the central role of the pineal gland in circadian regulation . This warrants further investigation of the influence of the pineal gland-intrinsic clock on circadian rhythms of behavior and other manifestations of circadian function . Here we address the role of the pineal gland in the zebrafish circadian system by generating and studying a novel transgenic line in which the molecular clock has been selectively blocked in the melatonin-producing cells of the pineal gland by a dominant-negative strategy . Characterization and analysis of this transgenic model establish the function of the molecular clock in the zebrafish pineal gland , probably as part of a multicomponent clock system , in regulating circadian rhythms of behavior .
To selectively block the core molecular circadian clock of the pineal gland we generated a transgenic zebrafish line , Tg ( aanat2:EGFP-ΔCLK ) ( Fig 1 ) , in which this clock is blocked by means of a dominant-negative strategy . This line expresses a C-terminal-truncated form of the zebrafish CLOCKa protein ( ΔCLK ) in the melatonin-producing cells of the pineal gland , under the control of the regulatory regions of the aanat2 gene [36] , which drive specific expression in these cells . The ΔCLK mutation was originally described in mice [37] and an equivalent mutation was later exploited to transiently block the clock in zebrafish embryos [38] . The ΔCLK protein consists of the bHLH and PAS domains , enabling it to heterodimerize with BMAL and bind to the E-box enhancer sequence , but it lacks the C-terminal glutamine-rich transactivation domain , thus abolishing its capacity to activate transcription . Therefore , ΔCLK displays a dominant-negative function by competing with endogenous CLOCK proteins [38] . The dominant-negative function of ΔCLK is evident in zebrafish Pac-2 cells , in which the expression of ΔCLK abolishes rhythmic expression driven by E-box elements ( S1 Fig ) . The Tg ( aanat2:EGFP-ΔCLK ) line also expresses enhanced green fluorescent protein ( EGFP ) , which facilitates the identification of positive transgenics and the dissection of ΔCLK-expressing pineal glands ( Fig 1B and 1B' ) . EGFP is separated from ΔCLK by the 2A peptide linker for production of two separate proteins [39] . Whole-mount immunostaining analysis confirms that Myc tag-labeled ΔCLK expression is restricted to the pineal gland ( Fig 1C ) . RNA-seq analysis indicates that ΔCLK is highly expressed in the adult Tg ( aanat2:EGFP-ΔCLK ) pineal gland ( 30-fold higher than the endogenous clock genes ) in a non-rhythmic manner ( chart C in S2 Fig ) . Advantages of this dominant-negative approach over gene knockout are tissue specificity and the ability to overcome possible gene redundancy resulting from the presence of multiple clock paralogs in zebrafish . Aanat2 encodes the enzyme that determines the rate of melatonin production in the pineal gland; a robust aanat2 mRNA rhythm starting at 2 days post-fertilization ( dpf ) is considered as a marker for circadian clock function [40 , 41] . To test whether the pineal gland molecular clock has been blocked in Tg ( aanat2:EGFP-ΔCLK ) fish , we first examined the expression of aanat2 in the larval pineal gland . Tg ( aanat2:EGFP-ΔCLK ) larvae and their wild-type ( WT ) siblings were entrained by seven LD cycles and then transferred to constant darkness ( DD ) . Larvae were collected at 4-hr intervals throughout one daily cycle and subjected to whole-mount in-situ hybridization ( ISH ) for aanat2 mRNA . A robust clock-controlled rhythm of aanat2 expression was observed in the pineal glands of WT sibling larvae; this rhythm , however , was absent in Tg ( aanat2:EGFP-ΔCLK ) larvae ( p<0 . 0001 , two-way ANOVA ) , which exhibited intermediate to high basal levels of aanat2 mRNA ( Fig 2 ) . Loss of the clock-controlled rhythm of aanat2 expression demonstrates that the molecular clock in the pineal melatonin-producing cells of Tg ( aanat2:EGFP-ΔCLK ) fish was effectively disrupted . To examine the effect of the ΔCLK mutation on rhythms of melatonin production , the pineal glands of Tg ( aanat2:EGFP-ΔCLK ) and control fish were cultured in a flow-through perfusion system . The lighting schedule was one dark-light ( DL ) cycle followed by two daily cycles under DD . The ΔCLK-expressing pineal glands produced a normal melatonin rhythm under the DL cycle , demonstrating that the ability to synthesize melatonin and light responsiveness were not affected by the mutation . The normal rhythmic secretion of melatonin observed in the ΔCLK-expressing pineal glands under the DL cycle reflects the effect of light on AANAT2 stability , resulting in the suppression of melatonin production . In contrast , under DD the melatonin rhythm was disrupted ( p<0 . 05 , Kolmogorov-Smirnov test; Fig 3 ) , characterized by increased basal levels of melatonin release during the subjective day . These results are in accordance with the expression pattern of aanat2 , and further validate the Tg ( aanat2:EGFP-ΔCLK ) line as a suitable model for studying the roles of the pineal clock and its primary output , melatonin , in driving circadian rhythms at the whole animal level . To further evaluate the effect of ΔCLK on the molecular clock in the pineal gland , we characterized circadian changes in the pineal gland transcriptome by means of mRNA-seq analysis . Pineal glands were sampled throughout two daily cycles under DD from adult Tg ( aanat2:EGFP-ΔCLK ) fish that were previously adapted to 24-hr LD cycles , replicating the experimental procedure formerly applied to Tg ( aanat2:EGFP ) fish by Tovin et al . ( Fig 4A; [42] ) . The data obtained from mRNA-seq were subjected to Fourier analysis and compared with the data from Tg ( aanat2:EGFP ) fish that served as controls ( Methods; [42] ) . Whereas 290 genes exhibited circadian rhythms of expression in control pineal glands , only 29 such genes were identified in Tg ( aanat2:EGFP-ΔCLK ) pineal glands ( Fig 4B; S1 and S2 Tables; false-detection rate , 10% ) . A set of 18 genes appeared in both lists , indicating that they maintained a circadian rhythmic profile in the ΔCLK-expressing pineal glands , albeit with reduced amplitudes . These included the core clock genes per1a , per1b , cry2a and cry3 , and the clock accessory loop genes reverbb2 , dec1 and dec2 . In general , the circadian profiles of core clock genes and clock accessory loop genes seemed to be only partially affected by the ΔCLK mutation ( S2 and S3 Figs ) . According to this analysis , 11 genes acquired a circadian rhythm of expression in the Tg ( aanat2:EGFP-ΔCLK ) pineal gland ( S4 Fig ) ; the regulatory mechanisms underlying the expression of these genes clearly require further examination . Importantly , the majority of CCGs lost their circadian profile in the Tg ( aanat2:EGFP-ΔCLK ) pineal gland ( Fig 4B ) , indicating that these genes are directly or indirectly regulated by the CLOCK/BMAL heterodimer , and that the output pathways of the pineal circadian clock are substantially impaired by the ΔCLK mutation . While some CCGs that became arrhythmic display intermediate or high overall expression levels compared with their expression in the Tg ( aanat2:EGFP ) pineal gland , the expression of others is down-regulated or completely abolished , and some CCGs maintained their circadian profiles in the Tg ( aanat2:EGFP-ΔCLK ) pineal gland ( Fig 5 ) , suggesting that CCGs are regulated by various mechanisms in addition to the molecular clock . The role of the pineal gland oscillator in coordinating peripheral molecular clocks was analyzed using a circadian clock-reporter zebrafish line , Tg ( −3 . 1 ) per1b::luc [43] , in which a luciferase reporter is expressed under the control of the per1b promoter , serving as a marker for peripheral clock rhythms . Tg ( aanat2:EGFP-ΔCLK ) fish were crossed with Tg ( −3 . 1 ) per1b::luc and the offspring , Tg[aanat2:EGFP-ΔCLK; ( −3 . 1 ) per1b::luc] and control Tg ( −3 . 1 ) per1b::luc larvae , were entrained under LD cycles and the luciferase activity in whole larvae was monitored under DD for two daily cycles . The pineal ΔCLK mutation did not significantly alter the reporter gene expression rhythms driven by the per1b promoter ( Fig 6 ) . This result implied that the pineal gland clock does not regulate peripheral molecular clocks , and is in accordance with the finding that peripheral clocks are not affected in melatonin-deficient fish [31] . This supports the hypothesis that peripheral clocks in zebrafish are independent of the pineal gland clock . However , this hypothesis should be taken with caution , because responses of small populations of peripheral clock-containing cells to melatonin may have been masked by the whole-larvae measurements , and because the activity of the per1b promoter may not be representative of the entire array of rhythmic genes in peripheral clock tissues . To determine the role of the pineal gland clock in generating clock-regulated behavioral rhythms , we analyzed the rhythmic locomotor activity of Tg ( aanat2:EGFP-ΔCLK ) and control larvae under various photic conditions . Zebrafish larvae exhibit daily rhythms of locomotor activity under constant conditions with higher levels of activity during the subjective day , determined by prior LD cycles [44] or by a single light pulse [45] . Under LD cycles , ΔCLK expression in the pineal gland had no effect on daily rhythms of locomotor activity ( Fig 7D; periods of 23 . 8±0 . 04 hr and 23 . 7±0 . 2 hr and amplitudes of 45 . 7±4 . 3 and 40 . 5±3 for control and ΔCLK larvae , respectively ) , reflecting the masking effects of light and dark . The masking effects of light and dark on clock-regulated locomotor activity of zebrafish larvae were further demonstrated under a 3 . 5-hr light: 3 . 5-hr dark schedule ( S5 Fig ) , in which larval activity was predominantly determined by the lighting conditions . However , when LD-entrained larvae were placed under DD ( Fig 7A ) or constant dim light ( DimDim; Fig 7B ) , circadian rhythms of locomotor activity were significantly affected in ΔCLK fish ( p<0 . 0001 , Kolmogorov-Smirnov test ) . Under DD , the period did not change significantly ( 24 . 9±0 . 6 hr and 23 . 7±0 . 8 hr for control and ΔCLK larvae , respectively ) , but the amplitude was significantly reduced in ΔCLK larvae ( 13 . 4±3 . 5 and 4 . 5±1 . 1 for control and ΔCLK larvae , respectively , p<0 . 05 , t-test ) . Likewise , under DimDim both groups exhibited a similar period ( 23 . 6±0 . 4 hr and 24±0 . 5 hr for control and ΔCLK larvae , respectively ) , but the amplitude was significantly reduced in ΔCLK larvae ( 34 . 7±4 . 5 and 17±2 . 7 for control and ΔCLK larvae , respectively , p<0 . 01 , t-test ) . Compared with DD , the use of DimDim produces higher levels of locomotor activity , higher amplitude rhythms and smaller variations in both control and Tg ( aanat2:EGFP-ΔCLK ) larvae . The residual rhythms of locomotor activity under DD or DimDim suggest that the pineal gland clock is not the sole regulator of these rhythms , and that additional clock centers , likely to be located in the central nervous system , contribute to their generation . Interestingly , under constant light ( LL ) , the locomotor activity rhythms were not affected by the blocked pineal clock ( Fig 7C; periods of 25 . 8±0 . 3 hr for both groups and amplitudes of 17 . 7±1 . 8 and 19 . 3±2 . 1 for control and ΔCLK larvae , respectively ) . Given that melatonin production is inhibited by light exposure and hence no melatonin is expected in either Tg ( aanat2:EGFP-ΔCLK ) or control larvae under LL , similar rhythms under these conditions hint at a possible role for melatonin in mediating the effects of the pineal gland clock on the measured behavioral outputs . Sleep in zebrafish larvae has been defined as an inactive bout of more than 1 minute , based on the finding that larvae exhibit reduced responsiveness after 1 minute of inactivity [46] . Sleep time analysis according to this criterion resulted in effects corresponding to those of the locomotor activity analysis: The day/night sleep time differences in ΔCLK larvae were disrupted under both DD and DimDim , but not under LL or LD cycles ( S6 Fig ) . To determine whether the reduced locomotor activity rhythms of ΔCLK larvae under DD or DimDim ( Fig 7A and 7B ) merely reflect increased sleep , we analyzed the activity of larvae during periods of wakefulness ( waking activity; [47] ) . This analysis revealed that the day/night differences in waking activity are markedly reduced in ΔCLK larvae under both DD and DimDim but not under LL or LD cycles ( S7 Fig ) . These results indicate that the effect of pineal ΔCLK expression on rhythms of locomotor activity does not simply reflect enhanced sleep . During sleep , animals show a preference for particular environmental locations and adopt distinctive postures . Adult zebrafish show a preference for either the top or the bottom of the tank during behavioral sleep [48] . Larval zebrafish tend to swim in the top third of the water column , rapidly descend toward the bottom of the tank upon loss of illumination [49] , and remain near the bottom of the tank during behavioral sleep [33] . However , it has not been previously determined whether these changes of posture in larvae are attributable to the lighting conditions or reflect an internal circadian state . We therefore entrained larvae with LD cycles and then monitored their position in the water column for 48 hr under DD . The larvae tended to swim in the top third of the water column . However , there was a robust daily rhythm in the magnitude of this preference , with a greater proportion of larvae found in the top zone during subjective day than during subjective night ( S8 Fig ) . Thus , as in adult zebrafish , place preference in larvae appears to be regulated by the circadian clock . To determine whether this rhythmic behavior is driven by the pineal gland circadian clock , Tg ( aanat2:EGFP-ΔCLK ) and control larvae were entrained by LD cycles and their position in the water column was monitored under DD for two daily cycles . As expected , control larvae displayed a clear clock-controlled rhythm of place preference , with significantly more larvae in the top third of the water column during the subjective day; however , no day/night differences in place preference were observed in Tg ( aanat2:EGFP-ΔCLK ) larvae ( Fig 8 ) . Thus , the circadian clock within the pineal gland not only modulates rhythms of locomotor activity but also drives circadian changes in place preference , suggesting that it functions as a clock center , contributing to behavioral rhythms in fish .
The organization of circadian systems is a topic of growing interest within the field of chronobiology . It is now apparent that there are considerable differences in how multiple clocks are organized into a network , how they are synchronized , and how they impact on physiology and behavior . The organization of the circadian system in mammals is considered to be hierarchical , with the SCN being the master clock that drives essentially all circadian output rhythms [8] . This organization differs markedly from the multicomponent system seen in birds , where multiple clock centers located in the pineal gland , retina , and hypothalamus operate coordinately to regulate physiological and behavioral rhythms [10] . An important common feature of clock centers is that they can be entrained by light . The current report addresses the role of a presumed clock center in a vertebrate species possessing directly light-entrainable peripheral clocks . As discussed below , this study of a unique model , in which the clock is selectively blocked in the melatonin-producing photoreceptor cells of the pineal gland , has provided evidence that this light-sensitive clock plays an important role in the zebrafish circadian system , as it augments behavioral rhythms . Thus , the pineal gland appears to serve as a clock center for behavior; however , it is not a master oscillator like the mammalian SCN , but is rather part of a system composed of multiple clock centers , as in the case of birds . Our findings indicate that expression of ΔCLK in the melatonin-producing cells of the pineal gland disrupts the clock-controlled rhythms of melatonin secretion ( Fig 3 ) , along with elimination of the clock-controlled mRNA rhythm of aanat2 in the larval pineal gland ( Fig 2 ) . Notably , the overall levels of both aanat2 expression and melatonin secretion remained intermediate to high in Tg ( aanat2:EGFP-ΔCLK ) fish , with increased subjective daytime levels compared with control fish . These results indicate that rhythms of aanat2 expression and melatonin production are predominantly driven by the core clock mechanism . An important effect of light on melatonin production is a rapid shutdown of its synthesis through a clock-independent mechanism , in which light promotes degradation of AANAT2 protein [50–53] . Accordingly , when placed under LD cycles , the melatonin secretion pattern appears normal in Tg ( aanat2:EGFP-ΔCLK ) pineal glands ( Fig 3 ) . In-depth analysis of the circadian transcriptome in ΔCLK-expressing pineal glands revealed that the majority of CCGs have lost their circadian expression profile ( Figs 4 and 5 ) , indicating that the molecular clock outputs have been blocked . On the other hand , the circadian expression patterns of core clock genes and of genes considered to form clock accessory loops were only partially disturbed ( S2 and S3 Figs ) . One possible explanation for the residual rhythmic expression of clock genes is that it derives from a functional molecular clock in neighboring pineal neuronal cells that do not produce melatonin and thus do not express ΔCLK . Another outcome of this analysis is the observation that the basal expression of the affected CCGs in the ΔCLK pineal gland ranges from being relatively high to being completely eliminated ( Fig 5 ) . This observation implies that the circadian clock regulates the rhythmic expression of CCGs , while in some cases additional factors contribute to driving their expression . Previous analysis of the zebrafish pineal circadian transcriptome showed that CCGs exhibit various phases of circadian expression [42] , a finding that also points to the contribution of additional factors and mechanisms in shaping their expression patterns . It is also likely that these additional factors interact with components of the circadian clock in the regulation of target gene expression . For example , the expression of zebrafish aanat2 was shown to be regulated by a synergistic interaction between the rhythmically expressed CLOCK/BMAL heterodimer and the photoreceptor-specific homeobox OTX5 [54] . As demonstrated by per1b:luc rhythms in intact larvae ( Fig 6 ) , the pineal clock and rhythmic melatonin production do not influence the global molecular rhythms of peripheral clocks . This is consistent with previous findings in melatonin-deficient larvae whose overall peripheral per3:luc rhythms remained unaltered [31] , suggesting that in zebrafish , peripheral molecular clocks function independently of the pineal clock . However , this hypothesis should be taken with caution , for two reasons . First , the measured activity of a single promoter may not be representative of all peripheral clock and clock-controlled genes . In mice , for example , the rhythmic expression of some genes in the liver is driven by the SCN , while other genes exhibiting circadian rhythms of expression are controlled by cell-autonomous clock mechanisms [55] . Secondly , measurements of peripheral clocks carried out at the level of the whole larva may have masked the contribution of small populations of peripheral clock-containing cells that are affected by pineal-derived signals , such as melatonin . One such example in mammals is the effect of melatonin on the rhythmic expression of clock and clock-controlled genes in the pars tuberalis through the MT1 receptor [56–58] . Our knowledge of the distribution of melatonin receptors in zebrafish is limited to three melatonin receptor-encoding mRNAs , mtnr1aa , mtnr1ba and mtnr1bb , which appear to be widely expressed in the embryonic brain [59] , two of which ( mtnr1ba and mtnr1bb ) are later restricted mainly to the periventricular gray zone of the optic tectum and the periventricular thalamus and hypothalamus of adult zebrafish [60] . However , the function and distribution of three other melatonin receptor-encoding genes ( mtnr1c , mtnr1al and mtnr1ab ) are still unknown . In other fish species , the expression of melatonin receptors has been shown to be widely distributed in the brain and extra-brain tissues [28 , 61] . Indeed , in zebrafish , exogenous melatonin was shown to gate cell proliferation during development [59] , inhibit injury-induced neutrophil migration [62] , modify gene expression in the liver and gonads [63 , 64] , and directly affect ovarian oocytes [63] . These reports indicate that melatonin exerts peripheral functions in zebrafish . It remains to be determined whether certain peripheral molecular rhythms in zebrafish are driven by melatonin . Our present findings confirm that vertical positioning of zebrafish larvae in a water column is a circadian clock-regulated behavior , with a preference toward the top part of the column during the subjective daytime ( S8 Fig ) . Under DD , circadian rhythms of positioning were absent in Tg ( aanat2:EGFP-ΔCLK ) larvae ( Fig 8 ) , indicating that this behavioral rhythm is driven by the pineal clock , thereby providing evidence for the function of the pineal gland as a behavior-regulating clock center . Clock-regulated rhythms of locomotor activity under DD and DimDim were also significantly affected by blocking the pineal oscillator , but were not eliminated; amplitudes were significantly reduced but period length remained unaltered ( Fig 7 ) . These findings demonstrate that while the pineal clock is fundamental to augmenting activity rhythms , additional clock centers are highly likely to play a role in driving the full repertoire of circadian behaviors . The finding that circadian control of melatonin production is absent in Tg ( aanat2:EGFP-ΔCLK ) fish is consistent with the view that in most non-mammalian vertebrates melatonin rhythms are driven by a pineal-intrinsic clock . The results of the present study support the hypothesis that melatonin is the link between the pineal clock and behavior . Locomotor activity rhythms and the corresponding sleep patterns were altered under DD and DimDim when melatonin is continuously present in Tg ( aanat2:EGFP-ΔCLK ) fish , in contrast to the rhythmicity observed in WT fish . Furthermore , this did not occur under LL , when neither of the strains produce melatonin , nor under LD cycles , when melatonin rhythms are similar in both strains ( Fig 7 ) . Similarly , decreased amplitudes of rhythmic locomotor activity were observed under DD in melatonin-deficient larvae [31] , supporting the view that melatonin contributes to the regulation of these rhythms in zebrafish . The current genetic evidence for involvement of the pineal clock and its melatonin output in driving rhythms of locomotor activity and of place preference in zebrafish is supported by classical earlier studies in various fish species [28] . Nevertheless , the zebrafish pineal gland may have additional hormonal or neuronal clock outputs that regulate circadian behavior , which have yet to be investigated . An important role for melatonin in zebrafish is in the induction of sleep . This was shown both pharmacologically by exogenous administration of melatonin [32 , 33] , and genetically by investigation of a melatonin-deficient zebrafish line [31] . Therefore , the attenuated locomotor activity rhythms measured under DD and DimDim in ΔCLK larvae may reflect the induction of sleep by the continuously high basal levels of melatonin under these conditions . However , pineal ΔCLK expression also affected locomotor activity during periods of wakefulness ( S7 Fig ) , suggesting that the pineal clock and possibly melatonin regulate activity rhythms independently of the somnogenic effect of melatonin . As shown in this study , the use of a strategy in which ΔCLK is selectively expressed in a target tissue makes it possible to selectively block clock outputs while leaving the target tissue intact . This method also avoids problems encountered when interpreting gene knockout experiments , such as the existence of gene redundancy and protein moonlighting ( where the targeted gene has additional , sometimes unknown , functions ) . This approach can potentially be used to selectively block the clock outputs in any tissue or cell type in order to study the function of specific peripheral clocks , and to further explore the organization of the circadian system and the generation of circadian behaviors in the intact animal . The nature of the additional clocks that contribute to the generation of circadian behavior still remains to be elucidated . One approach would be to genetically block the circadian clock in other specific cell subsets in the brain by the generation of additional transgenic lines that express ΔCLK under the control of various tissue-specific enhancers . These could include , for example , hypocretin neurons in the hypothalamus that have been implicated in the regulation of sleep/wake transitions in zebrafish [65] , deep brain photoreceptor-expressing neurons [49] , and projection neurons of the pineal gland . Thus , use of this genetic approach to study the functional significance of neuron-intrinsic circadian clocks and their importance for neuronal activity and the physiology and behavior of the intact organism can be expected to have a broad impact on the field of neurobiology .
All procedures were approved by the Tel-Aviv University Animal Care Committee ( L-10-011 and L-15-047 ) and conducted in accordance with the National Council for Animal Experimentation , Ministry of Health , Israel . The pT2-aanat2:EGFP-2A-Δclocka-5×MYC construct ( Fig 1A ) was generated using the backbone of the Tol2 transposable element-containing vector , pT2KXIGΔin [66] . The coding sequences of EGFP and a truncated zebrafish CLOCKa ( ΔCLK; [38] ) linked to 5 Myc tag sequences were cloned downstream of the aanat2 regulatory regions [36] . A sequence encoding the 'self-cleaving' 2A peptide was cloned between the open reading frames of EGFP and ΔCLK , in order to produce two distinct proteins [39] . The sequence of the pT2-aanat2:EGFP-2A-Δclocka-5×MYC construct is provided in S9 Fig . The transgenic line , Tg ( aanat2:EGFP-ΔCLK ) , registered in the Zebrafish Model Organism Database ( ZFIN ) as Tg ( aanat2:EGFP-2A-Δclocka-5×MYC ) tlv03 , was generated using the Tol2 system as described [67] . Tol2 plasmids were kindly provided by Koichi Kawakami . Capped Tol2 transposase mRNA was synthesized in vitro using the mMESSAGE mMACHINE SP6 Transcription Kit ( Ambion ) and a linearized pCS-TP plasmid as template . Approximately 1 nl of a DNA/RNA solution containing 25 ng/μl of the pT2-aanat2:EGFP-2A-Δclocka-5×MYC circular DNA and 25 ng/μl of transposase mRNA were injected into fertilized eggs at the single-cell stage . Founder ( F0 ) fish were raised to adulthood and outcrossed to screen for integration of the transgene into the germline . Transgenic EGFP-expressing progeny ( F1 ) were isolated using a fluorescence stereomicroscope . Several transgenic lines were obtained , all of them expressing EGFP specifically in the melatonin-producing pineal cells . F1 fish that showed strong transgene expression were further propagated . F2 progeny from one selected outcrossed F1 fish were further incrossed to generate homozygotes in the F3 generation . F3 homozygotes were further incrossed to produce homozygous transgenic progeny . WT siblings of F3 homozygotes were also maintained and further incrossed to produce control WT progeny . As in the case of transient expression of ΔCLK in embryos [38] , the stable expression of ΔCLK in the pineal gland did not affect overall embryonic development . Tg ( aanat2:EGFP-ΔCLK ) larvae ( 5 dpf ) were fixed , and whole-mount immunostaining was carried out as previously described [68] with mouse anti-Myc antibody 1:100 ( clone 9E10; Santa Cruz Biotechnology ) . Anesthetized or immunostained Tg ( aanat2:EGFP-ΔCLK ) larvae were placed in low melting point agarose . Images were obtained using a Leica TCS SP8 confocal laser scanning microscope equipped with Leica LAS AF image acquisition software . Heterozygous Tg ( aanat2:EGFP-ΔCLK ) larvae and their WT siblings were entrained by seven 12-hr:12-hr LD cycles , transferred to DD and sampled at 4-hr intervals throughout one daily cycle . The collected embryos were fixed and subjected to whole-mount ISH followed by quantification as previously described [69 , 70] , using the aanat2 probe [40] . Adult fish were raised in a temperature-controlled recirculation water system under 12-hr:12-hr LD cycles and their pineal glands were collected and cultured in a flow-through system as previously described [53] . Fish were anesthetized in 1 . 5 mM Tricane ( Sigma-Aldrich ) , decapitated , and pineal glands were removed by a surgical procedure carried out under a dissecting microscope ( Olympus SZX12 ) . Three pineal glands were collected from homozygous Tg ( aanat2:EGFP-ΔCLK ) fish , and three from control fish ( progeny of WT siblings ) . The pineal glands were kept in culture medium and placed in a flow-through system . Each individual gland was placed in a glass column and continuously perfused with medium ( 1 ml/hr ) delivered by a multi-channel peristaltic pump ( Minipuls 3; Gilson ) . The culture medium was MEM ( Sigma-Aldrich ) , supplemented with 2 mM L-glutamine , 0 . 1 mM L-tryptophan , 0 . 02 M sodium bicarbonate , penicillin ( 100 , 000 U/l ) –streptomycin ( 100 mg/l ) ( Biological Industries ) and 2 . 5 mg/l Fungizone ( Amphotericin B; Sigma-Aldrich ) . The medium was continuously bubbled with a 5% CO2: 95% O2 gas mixture during the experiment . Fractions ( 1 ml ) of medium were collected at 1-hr intervals using a multi-channel fraction collector ( FC204; Gilson ) . The apparatus was placed inside a light- and temperature-controlled incubator and the temperature was maintained at 24°C . Pineal glands were exposed to one DL cycle followed by two DD cycles . The concentration of melatonin in the collected medium was determined by high performance liquid chromatography ( HPLC ) using a 125 × 4 . 6 mm C18 ( 2 ) reversed-phase analytic column ( Phenomenex Luna ) with a particle size of 5 μm and a Dionex UltiMate 3100 fluorescence detector ( Thermo Scientific ) . A volume of 100 μl was injected from samples at 2-hr intervals . The excitation and emission wavelengths were 280 nm and 340 nm , respectively . The mobile phase consisted of 0 . 1 M Na2HPO4 containing 20% acetonitrile; the pH was adjusted to 6 . 5 with orthophosphoric acid . The mobile phase flow was 1 . 5 ml/min and the melatonin retention time ( about 7 min ) was confirmed using a commercial standard ( Sigma-Aldrich ) . To account for variation in the basal levels of melatonin secretion from individual pineal glands , melatonin levels secreted by each pineal gland were normalized by dividing the absolute levels by the maximal night-time levels . Levels of melatonin production by each pineal gland under DD underwent Fourier analysis and were scored with a G-factor ratio to determine circadian rhythmicity ( see 'Fourier analysis' in S1 Text ) . Statistical differences between Tg ( aanat2:EGFP-ΔCLK ) and control fish in rhythmic melatonin production under DD were determined by the Kolmogorov-Smirnov test . Adult homozygous Tg ( aanat2:EGFP-ΔCLK ) fish were raised in a temperature-controlled recirculation water system under 12-hr:12-hr LD cycles , and transferred to DD at the end of the light period prior to sampling . Pineal glands were sampled at 4-hr intervals throughout two daily cycles under DD at 12 time points corresponding to circadian time ( CT ) 14 , 18 , 22 , 2 , 6 , 10 , 14b , 18b , 22b , 2b , 6b and 10b ( Fig 4A ) , as previously described [42] . A pool of 16 pineal glands was collected at each time point . In addition , two control pools of 14 pineal glands were collected from Tg ( aanat2:EGFP ) fish at time points corresponding to CT2 and CT14b . Fish were anesthetized in 1 . 5 mM Tricane ( Sigma-Aldrich ) and decapitated . Fluorescent pineal glands were selectively removed under a dissecting microscope ( Olympus SZX12 ) equipped with filters for excitation ( 460–490 nm ) and emission ( 510–550 nm ) of EGFP . Total RNA for mRNA analysis was isolated using RNeasy Lipid Tissue Mini Kit ( Qiagen ) . mRNA-seq data acquisition and analysis were carried out as a replicate of a previously described procedure [42] . The Illumina TruSeq protocol was used to prepare libraries from RNA samples . Overall , 14 libraries [12 time points from Tg ( aanat2:EGFP-ΔCLK ) fish and two control samples from Tg ( aanat2:EGFP ) fish] were run on a single flow cell of an Illumina HiSeq2500 machine ( rapid run mode ) using the multiplexing strategy of the TruSeq protocol ( Institute of Applied Genomics , Italy ) . On average , 14 million paired-end reads were obtained for each library . The reads were of 2×50 base pairs . The sequencing data were deposited in the Sequence Read Archive , under accession SRP016132 . TopHat [71] was used for aligning the reads against the zebrafish genome , keeping only uniquely aligned reads with up to two mismatches per read . On average , 68% of the reads had unique alignment with the zebrafish genome . Reads aligned with the protein coding regions of known NCBI reference sequence ( RefSeq ) genes were used . A custom script written in Perl was used to parse the output of TopHat , which is given in Sequence Alignment/Map ( SAM ) format ( http://samtools . sourceforge . net/ ) , and to convert it into a raw number of reads aligned to each position in each RefSeq gene . The RefSeq genes data was obtained from the Table Browser of the UCSC genome browser ( genome . ucsc . edu/ ) using the zebrafish July 2010 ( Zv9/danRer7 ) assembly . To avoid PCR duplicates , only paired-end reads with unique start positions in the genome in both pairs were used [72] . The 26 mRNA-seq profiles ( i . e . , the number of reads aligned against each RefSeq gene ) corresponding to the 12 time points using pineal glands of Tg ( aanat2:EGFP-ΔCLK ) fish , the same 12 time points using pineal glands of control Tg ( aanat2:EGFP ) fish ( data from Tovin et al . [42] ) , and the two control pineal gland samples from Tg ( aanat2:EGFP ) fish ( CT2 and CT14b ) were normalized together . The logarithmically transformed dataset was normalized using quantile normalization [73] . Transcripts with low maximum expression values ( i . e . , their highest level over all time points is in the lower quartile of all transcripts ) were not included in the Fourier analysis . Fourier analysis was conducted as previously described ( [42]; S1 Text ) . See S1 Text . Tg ( aanat2:EGFP-ΔCLK ) fish were crossed with Tg ( −3 . 1 ) per1b::luc fish [43] . Embryos were raised under 12-hr:12-hr LD cycles at 25°C . At 3 dpf , single larvae were transferred into individual wells of a 96-multiwell plate ( Nunc ) in E3 media ( without methylene blue ) supplemented with 0 . 5 mM beetle luciferin potassium salt solution ( Promega ) , and the plate was sealed using an adhesive TopSeal sheet ( Packard ) . Plates were then subjected to two LD cycles , followed by two daily cycles under DD , under which bioluminescence from whole larvae was assayed using a TopCount NXT Scintillation Counter ( 2-detector model; Packard ) . Bioluminescence data were analyzed using the Import and Analysis Macro ( I&A , Plautz and Kay , Scripps ) for Microsoft Excel or CHRONO software [74] . Short-term and long-term trends were removed from the raw data by an adjacent-averaging method with 3-hr and 2-day running means , respectively . Normalized values were obtained by dividing the bioluminescence values by their average value for each larva . Traces represent the mean value ± SD of 23 Tg[aanat2:EGFP-ΔCLK; ( −3 . 1 ) per1b::luc] larvae and 55 control Tg ( −3 . 1 ) per1b::luc larvae; each group was comprised of larvae from two separate crosses . To determine the circadian rhythmicity of per1b promotor activity , bioluminescence data from each larva underwent Fourier analysis and were scored with a G-factor ratio ( see 'Fourier analysis' in S1 Text ) . Kolmogorov-Smirnov test was applied to compare the distribution of G-factors between Tg[aanat2:EGFP-ΔCLK; ( −3 . 1 ) per1b::luc] and control Tg ( −3 . 1 ) per1b::luc larvae . See S1 Text . Homozygous Tg ( aanat2:EGFP-ΔCLK ) embryos and control embryos ( progeny of WT siblings ) were raised in a light- and temperature-controlled incubator under 12-hr:12-hr LD cycles at 28°C . On the 4th day of development , larvae were placed in 48-well plates in the observation chamber of the DanioVision tracking system ( Noldus Information Technology ) for acclimation under controlled temperature ( 28°C ) and lighting conditions ( LED; intensity of 'light' and 'dim light' were 1 . 8 W/m2 and 0 . 013 W/m2 , respectively ) according to the desired protocol . Starting from the 6th day of development , movement was tracked and analyzed by the Ethovision 11 . 0 software ( Noldus Information Technology ) . For the analysis of locomotor activity , the raw data were converted into the total distance moved ( cm ) by each larva per 10 min time-bins . The data are presented as a moving average ( 20 sliding points ) of 24 larvae in each group , excluding experiments with light-dark or dark-light transitions that trigger a temporary rise in activity ( Fig 7D and S4 Fig ) . To determine alterations in circadian rhythms of locomotor activity , individual tracks underwent Fourier analysis and were scored with a G-factor ratio ( see 'Fourier analysis' in S1 Text ) . Differences in G-factor distributions between Tg ( aanat2:EGFP-ΔCLK ) and control groups were determined by the Kolmogorov-Smirnov test . The periods of locomotor activity rhythms were computed by the chi-square periodogram [75] with ActogramJ software ( [76]; http://actogramj . neurofly . de/ ) , and statistical differences between Tg ( aanat2:EGFP-ΔCLK ) and control larvae were determined by t-test . Amplitude values were calculated as the difference between the peak of activity at day 7 and the preceding trough , divided by 2 , and statistical differences between Tg ( aanat2:EGFP-ΔCLK ) and control larvae were determined by t-test . For analyzing the experiment in which activity was monitored under 3 . 5-hr light: 3 . 5-hr dark cycles ( masking protocol , S5 Fig ) , the percentage of activity during the bouts of light and dark was calculated , and the difference between Tg ( aanat2:EGFP-ΔCLK ) and control larvae was determined by t-test . For the analysis of sleep and waking activity , the raw data were converted to the number of seconds spent moving per 1-min time bin for each larva , with stop velocity threshold of 0 . 59 cm/s and start velocity threshold of 0 . 6 cm/s [65] . Sleep time was calculated as the number of minutes without movement per 1-hr . Waking activity was computed as the average number of seconds of activity/waking minutes per 1-hr . Data are presented as the average sleep time or average waking activity of 24 larvae in each group , and also as the average sleep time or average waking activity during subjective day and subjective night periods . Repeated-measures ANOVA was applied to compare the sleep time and log-transformed waking activity values of Tg ( aanat2:EGFP-ΔCLK ) and control groups during subjective day and subjective night periods . For place preference , larvae were tested in groups of 15 in chambers of 15 mm × 22 mm × 50 mm ( width , length , height ) . Larvae were initially entrained under LD cycles for 4 days , and were then placed in the chamber under DD . Recordings were initiated after a 24-hr acclimation period , using a μEye IDS-1545LE-M CMOS camera ( 1stVision ) to capture a snapshot of each group every 15 sec . Backlit illumination was provided by an 880-nm infrared LED array ( Advanced Illumination ) which was activated for 50 ms , synchronized with image acquisition . Timing was controlled by DAQtimer event control software and image analysis to detect the position of larvae performed using FLOTE [77] . For analysis , we then calculated the proportion of larvae in each of three identically sized zones of the chamber: top , middle and bottom . For behavioral testing we used homozygous Tg ( aanat2:EGFP-ΔCLK ) larvae , and progeny of WT siblings as controls . The raw data were converted into the average percentage of larvae in the top third of the water column per 5-min time-bins . The data are presented as a moving average ( 40 sliding points ) of four groups of Tg ( aanat2:EGFP-ΔCLK ) larvae and eight groups of control larvae . To determine alterations in circadian rhythms of environmental positioning , the data underwent Fourier analysis and were scored with a G-factor ratio ( see 'Fourier analysis' in S1 Text ) . Differences in the G-factor distributions between Tg ( aanat2:EGFP-ΔCLK ) and control groups were determined by the Kolmogorov-Smirnov test . Differences in the average percentage of larvae in the top third of the water column between CT 5–6 ( subjective day ) and CT 17–18 ( subjective night ) , under two daily cycles , were determined by paired t-test .
|
Most physiological and behavioral processes exhibit daily rhythms which are driven by an internal timing mechanism known as the circadian clock . Circadian systems are thought to be organized in a hierarchical manner , with a central pacemaker in the brain regulating peripheral clocks throughout the body . In fish species , the pineal gland has been considered to function as a central circadian clock organ . Nevertheless , a central role for the pineal gland in the zebrafish circadian system has been questioned , because peripheral clocks in this species are independently synchronized by light . Here we developed a genetically modified zebrafish model in which the molecular clock is selectively blocked in the melatonin-producing cells of the pineal gland . As a result , clock-controlled melatonin production and gene expression in the pineal gland are interrupted . Although the independent peripheral clocks are not affected by this genetic manipulation , circadian rhythms of behavior are attenuated . These findings indicate that the zebrafish pineal gland clock contributes to the generation of daily behavioral rhythms , possibly as part of a multiple pacemaker system .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"pineal",
"gland",
"vertebrates",
"animals",
"biomechanics",
"biological",
"locomotion",
"hormones",
"circadian",
"oscillators",
"animal",
"models",
"osteichthyes",
"developmental",
"biology",
"model",
"organisms",
"chronobiology",
"research",
"and",
"analysis",
"methods",
"fishes",
"life",
"cycles",
"circadian",
"rhythms",
"biochemistry",
"zebrafish",
"melatonin",
"anatomy",
"physiology",
"endocrine",
"system",
"biology",
"and",
"life",
"sciences",
"larvae",
"organisms"
] |
2016
|
Genetically Blocking the Zebrafish Pineal Clock Affects Circadian Behavior
|
In vivo fluorescence microscopy and electron cryo-tomography have revealed that chemoreceptors self-assemble into extended honeycomb lattices of chemoreceptor trimers with a well-defined relative orientation of trimers . The signaling response of the observed chemoreceptor lattices is remarkable for its extreme sensitivity , which relies crucially on cooperative interactions among chemoreceptor trimers . In common with other membrane proteins , chemoreceptor trimers are expected to deform the surrounding lipid bilayer , inducing membrane-mediated anisotropic interactions between neighboring trimers . Here we introduce a biophysical model of bilayer-chemoreceptor interactions , which allows us to quantify the role of membrane-mediated interactions in the assembly and architecture of chemoreceptor lattices . We find that , even in the absence of direct protein-protein interactions , membrane-mediated interactions can yield assembly of chemoreceptor lattices at very dilute trimer concentrations . The model correctly predicts the observed honeycomb architecture of chemoreceptor lattices as well as the observed relative orientation of chemoreceptor trimers , suggests a series of “gateway” states for chemoreceptor lattice assembly , and provides a simple mechanism for the localization of large chemoreceptor lattices to the cell poles . Our model of bilayer-chemoreceptor interactions also helps to explain the observed dependence of chemotactic signaling on lipid bilayer properties . Finally , we consider the possibility that membrane-mediated interactions might contribute to cooperativity among neighboring chemoreceptor trimers .
The chemotaxis signal transduction pathway [1] allows bacteria to respond to minute relative changes in chemical concentration over several orders of magnitude in ambient chemical concentration [2] , and ranks among the most studied signaling pathways in biology . The extreme sensitivity of the chemotaxis system results from amplification of external signals coupled with adaptation to persistent stimuli [3]–[7] . Fluorescence resonance energy transfer ( FRET ) experiments [8] , [9] have revealed that a crucial step in signal amplification occurs at the level of chemoreceptors: chemoreceptors signal in cooperative teams [8]–[10] . Indeed , Ising [11]–[13] and Monod-Wyman-Changeux [9] , [14]–[16] models of coupled teams of chemoreceptors both achieve quantitative agreement with the FRET data . The fundamental assumption underlying these models of signaling teams is that chemoreceptors do not respond independently to changes in the external ligand concentration , but rather each receptor influences the collective state of a team of neighboring receptors . Thus , the observed functional characteristics of chemotactic signaling rely on cooperative local interactions among chemoreceptors , and suggest a well-defined spatial organization of chemoreceptors . From a structural perspective , chemoreceptors are homodimers , which interact strongly to form trimers-of-dimers [17] , [18] . Recent breakthroughs in in vivo electron cryo-tomography have revealed [19]–[25] that chemoreceptor trimers form two-dimensional honeycomb lattices in which each trimer has three nearest-neighbors arranged in a face-on orientation . The honeycomb lattice architecture and characteristic lattice constant of 12 nm appear to be universally conserved among bacterial species [22] . Functional complexes require chemoreceptors plus the linker/kinase CheA and the linker protein CheW [19] , [20] , which may mediate cooperative interactions among neighboring trimers [9] , [10] . Fluorescence experiments have indicated that chemoreceptor lattices can exhibit variable stoichiometries of chemoreceptors , CheA , and CheW [9] , [26] , [27] , while clustering of chemoreceptors requires neither CheA nor CheW [27] , [28] . The size of chemoreceptor clusters can range from tens to thousands of receptors , with large chemoreceptor clusters observed predominantly at the cell poles but smaller clusters also found in the midcell regions [21]–[25] , [27]–[30] . Superresolution light microscopy of chemoreceptor lattices has suggested a stochastic model for cluster assembly [28] , [30] in which self-assembly of chemoreceptor lattices proceeds by nucleation and growth . Such stochastic self-assembly of chemoreceptor lattices relies on the existence of attractive interactions between chemoreceptor trimers , but in principle does not require direct cytoskeletal involvement or active transport of chemoreceptors . What are the molecular mechanisms yielding attraction between chemoreceptor trimers and , hence , self-assembly of chemoreceptor lattices ? Chemoreceptors are transmembrane proteins localized in the cytoplasmic membrane of bacteria . In general , membrane proteins deform the surrounding lipid bilayer [31] , [32] , which can lead to membrane-mediated interactions between proteins [33] , [34] . Thus , while chemoreceptors can be coupled by protein-protein interactions [9] , [10] , they may also interact via the cytoplasmic membrane . Here we develop a biophysical model of membrane-mediated interactions between chemoreceptor trimers which shows that membrane-mediated interactions can yield stochastic cluster assembly even at very dilute trimer concentrations . The model correctly predicts the observed face-on orientation of chemoreceptor trimers at small trimer separations [19] , [20] and suggests a series of “gateway” states for chemoreceptor lattice assembly . We find that the three-fold-symmetric directionality of membrane-mediated interactions between trimers can stabilize the observed honeycomb architecture of chemoreceptor lattices [19] , [20] even at suboptimal stoichiometries of chemoreceptors , CheA , and CheW [9] , [26] , [27] . The model also suggests a simple mechanism by which bilayer-chemoreceptor interactions can localize large chemoreceptor clusters to the cell poles even in the absence of interactions with CheA and CheW [27] . Furthermore , based on the assumption that the chemotactic signaling state impacts the hydrophobic thickness of chemoreceptors , our model allows us to quantify the membrane contribution to chemotactic signaling . In agreement with previous experimental observations [35]–[37] we find a dependence of chemotactic signaling on lipid bilayer properties . Finally , we examine the possibility of membrane-mediated cooperative signaling among neighboring chemoreceptor trimers .
In our analysis of membrane-mediated interactions between chemoreceptor trimers we follow the standard membrane-mechanical framework [31]–[33] for describing bilayer-protein interactions , and model chemoreceptor trimers as rigid membrane inclusions inducing elastic deformations in the surrounding lipid bilayer membrane . Such deformations can take the form of thickness deformations ( Fig . 1A ) , which originate from a hydrophobic thickness mismatch between chemoreceptors and the lipid bilayer , and midplane ( curvature ) deformations ( S1 Figure ) , which may be induced by a conical shape of chemoreceptor trimers resulting from a tilt in the transmembrane helices . To leading order , the elastic energies associated with thickness and midplane deformations decouple from each other , and can therefore be analyzed separately ( see S1 Text section 1 ) . We focus here on bilayer-chemoreceptor interactions and , hence , only consider the transmembrane regions of trimers in our model , with the peri- and cytoplasmic regions of trimers in Fig . 1A and S1 Figure only being shown for illustration . In general , neighboring membrane proteins are expected to induce overlapping deformation fields of the bilayer membrane , yielding [33] , [34] membrane-mediated interactions between proteins . Thus , membrane proteins can interact over several nanometers [33] , [34] , [38] without being in direct protein-protein contact . Membrane-mediated interactions due to thickness/curvature deformations induced by identical proteins are generally expected [33] to be attractive/repulsive at small protein separations , with thickness deformations yielding stronger membrane-mediated interactions than curvature deformations . Based on previous work concerning the far-field limit of membrane-mediated interactions between conical membrane inclusions [39] , [40] , analytic series solutions describing membrane-mediated interactions between proteins of arbitrary symmetry and at arbitrary separation have recently been developed [41] . Here we employ these analytic series solutions to determine the membrane-mediated interactions between chemoreceptor trimers . We find that midplane interaction energies resulting from the conical shape of trimers are typically well below and more than an order of magnitude smaller than thickness interaction energies ( see S1 Text section 2 ) . We therefore focus on membrane-mediated interactions between chemoreceptor trimers due to thickness deformations ( Fig . 1A ) . Our biophysical model of membrane-mediated interactions between chemoreceptor trimers is based on the standard framework of membrane mechanics [42]–[46] . We represent the lipid bilayer within the Monge representation for curved surfaces using the functions and , which define the heights of the hydrophilic-hydrophobic interfaces at the coordinates along the outer and inner membrane leaflets . The thickness deformations correspond to ( 1 ) where is one-half the hydrophobic thickness of the unperturbed lipid bilayer . Following previous work on bilayer-protein interactions [31]–[33] , [47] we describe the energetic cost of thickness deformations by the functional ( 2 ) where is the bending rigidity , is the stiffness associated with thickness deformations , and is the membrane tension . The term in Eq . ( 2 ) captures the energetic cost of membrane bending , while the term provides a simple description of the energetic cost of compressing or expanding the lipid bilayer . Typical measured values of and are and nm [33] , [48] , which we used for all the calculations described here . The term describes the effect of membrane tension on membrane undulations [42]–[47] , [49] , [50] . For generality we also allow for the term in Eq . ( 2 ) [38] , [45] , [48] , which captures the effect of membrane tension on lipid surface area under conservation of lipid volume . Phenomenological membrane deformation energies of the form in Eq . ( 2 ) have been employed to describe protein-induced bilayer thickness deformations in a range of systems [31]–[33] , [38] , [46]–[55] , and can be systematically refined [41] , [56]–[68] to provide a more detailed model of bilayer-protein interactions . The thickness deformation energy in Eq . ( 2 ) scales approximately with the square [31]–[33] of the hydrophobic mismatch which , in turn , is equal to the difference between one-half the trimer hydrophobic thickness , , and . A typical value of for the E . coli cytoplasmic membrane is nm [69] while , for example , the approximate hydrophobic thickness of the chemoreceptor Trg is nm [70] . The resulting hydrophobic mismatch nm yields a thickness deformation energy of the order of for a single chemoreceptor trimer , which induces strong membrane-mediated interactions between neighboring trimers ( see the Results section ) . For a given value of , the value of and hence the magnitude and sign of the hydrophobic mismatch , can be tuned by changing the membrane composition which , as demonstrated for gramicidin [47] , [71] and mechanosensitive [49] , [72] channels , allows for direct experimental tests of membrane-mechanical models of bilayer-protein interactions . We study here membrane-mediated interactions between chemoreceptor trimers as a function of hydrophobic mismatch . Thus , while we use in our calculations a chemoreceptor hydrophobic thickness consistent with Trg , our conclusions can be applied equally to other chemoreceptors . In the absence of detailed structural information on the transmembrane region of chemoreceptor trimers , we adopt a highly simplified model designed to capture two key features of chemoreceptor trimers: ( 1 ) As described above , we assume that chemoreceptors have a hydrophobic mismatch with the lipid bilayer , which induces membrane-mediated interactions between neighboring chemoreceptor trimers ( Fig . 1A ) . ( 2 ) In addition , the characteristic three-fold symmetry of chemoreceptor trimers yields directionality in membrane-mediated interactions between chemoreceptor trimers ( Fig . 1B ) . In particular , different relative orientations of neighboring chemoreceptor trimers produce distinct deformations of the bilayer membrane , resulting in a dependence of the energy of membrane-mediated interactions on the relative trimer orientation . Thus , membrane-mediated interactions between chemoreceptor trimers not only depend on the hydrophobic mismatch between chemoreceptors and the bilayer membrane [33] , but also on the distinctive three-fold symmetry of chemoreceptor trimers . While the precise size of the transmembrane cross section of trimers is not crucial for our model predictions , we allow for a finite characteristic size of trimers and lipids , which imposes steric constraints on the minimum edge-to-edge separation of neighboring trimers . Our simple model for the shape of chemoreceptor trimers ( Fig . 1B ) is consistent with recent electron cryo-tomography studies [19] , [20] . However , we focus here on the effects of generic aspects of chemoreceptor trimers , such as their symmetry , on membrane-mediated interactions , and our predictions do not rely on the detailed supramolecular shape of trimers . In particular , data obtained from electron cryo-tomography [19] , [20] , [73] mostly pertains to the cytoplasmic regions of chemoreceptor trimers , and the transmembrane structure of chemoreceptor trimers remains unknown . Indeed , the chemoreceptor dimers forming a trimer may spread apart within the membrane [19] , [20] , [73] , with the lipid bilayer infiltrating chemoreceptor trimers . Such lipid-chemoreceptor complexes would imply membrane-mediated interactions between the chemoreceptor dimers forming a trimer . Here we do not consider membrane-mediated interactions within trimers and , instead , focus on membrane-mediated interactions between chemoreceptor trimers . Thus , our model of the transmembrane shape of chemoreceptor trimers ( Fig . 1B ) may correspond to chemoreceptor trimers composed of only proteins as well as lipid-chemoreceptor complexes . For simplicity , we assume a constant hydrophobic thickness of chemoreceptor trimers . More detailed descriptions would allow for a variation of the hydrophobic thickness along the trimer circumference , which may result from details of the transmembrane structure of chemoreceptors or the formation of lipid-chemoreceptor complexes . Multiple lines of evidence [74]–[78] have indicated that chemoreceptor dimers signal the binding of a ligand across the cytoplasmic membrane through a piston-like sliding of one of the four transmembrane helices relative to the other three helices , by approximately 0 . 16 nm . This suggests that chemotactic signaling perturbs the hydrophobic surface of chemoreceptors and , indeed , it has been found [35]–[37] that bilayer-chemoreceptor interactions affect chemotactic signaling . Furthermore , the in vivo signaling response of chemoreceptors implies [9] , [16] that trimers exhibit strong cooperativity , and are either in the fully active or the fully inactive state . We account for these observations by assuming that chemoreceptor trimers can be active or inactive , with active and inactive trimers exhibiting a difference in hydrophobic thickness . For simplicity , we also assume that this difference in hydrophobic thickness is uniform along the trimer circumference , and is approximately equal to 0 . 16 nm as indicated by the piston model of chemotactic signaling [74]–[78] . ( While consistent with the observed role of the membrane in chemotactic signaling , this working model is highly simplified; more detailed models would allow , for instance , for the possibility of a tilt in the transmembrane helices upon switching [74] , [79] , for variations in the shift in hydrophobic thickness along the trimer circumference , and for possible differences in the hydrophobic surfaces exhibited by distinct chemoreceptors . ) The predicted strength of the coupling between bilayer properties and chemotactic signaling depends on model details , but the basic mechanism for membrane-mediated cooperativity among chemoreceptors considered here relies only on a difference in hydrophobic thickness between active and inactive trimer states .
We followed the approach developed in Refs . [39]–[41] to obtain analytic expressions for the energy of membrane-mediated interactions between chemoreceptor trimers due to thickness deformations . The energy is a function of center-to-center distance between trimers , , relative trimer orientation , membrane tension , and hydrophobic mismatch ( see S1 Text section 1 ) . A negative value of the energy of membrane-mediated interactions , , implies energetically favorable interactions between chemoreceptor trimers . For a hydrophobic mismatch corresponding to chemoreceptors and the cytoplasmic membrane of E . coli , we find three regimes of membrane-mediated interactions between chemoreceptor trimers ( Fig . 2A ) : ( 1 ) For trimer separations greater than nm membrane-mediated interactions are negligible , yielding energies smaller than . ( 2 ) For intermediate trimer separations , from – nm ( depending on relative trimer orientation ) up to nm , interactions are weakly unfavorable . ( 3 ) For small trimer separations , smaller than – nm ( depending on relative trimer orientation ) , membrane-mediated interactions are strongly favorable . In particular , we find that for the smallest values of allowed by steric constraints on lipid size , corresponding to a minimum edge-to-edge separation between trimers of approximately 0 . 8 nm , membrane-mediated interactions can reduce the thickness deformation energy by more than 15 compared to noninteracting chemoreceptor trimers . The interaction potentials in Fig . 2A show that membrane-mediated interactions yield strong attraction between chemoreceptor trimers over several nanometers , which suggests that membrane-mediated interactions may be sufficient for nucleation and growth of chemoreceptor lattices . Indeed , while chemoreceptors interact with CheA and CheW to form ordered lattices [19] , [20] , clustering of chemoreceptors requires neither CheA nor CheW [27] , [28] . Furthermore , superresolution light microscopy of chemoreceptor clusters has suggested [28] , [30] that chemoreceptor lattices self-assemble by stochastic nucleation of small clusters and capture of diffusing receptors by preexisting clusters . Fig . 2A implies that membrane-mediated interactions provide a plausible biophysical mechanism for the efficient self-assembly of chemoreceptor lattices via stochastic nucleation and capture . In particular , based on the statistical mechanics of phase segregation [80] , [81] the interaction energies in Fig . 2A allow us to estimate the critical trimer concentration for nucleation and growth of chemoreceptor lattices in the E . coli cytoplasmic membrane ( see S1 Text section 3 ) . We find that the critical trimer concentration for clustering is already reached with approximately 15 chemoreceptor trimers in the cytoplasmic membrane . This means that , even if trimers are very dilute in the cytoplasmic membrane , membrane-mediated interactions can lead to nucleation and growth of chemoreceptor lattices . Our model predicts that chemoreceptor clustering due to membrane-mediated interactions shows a characteristic dependence on hydrophobic mismatch ( Fig . 2B ) and membrane tension ( Fig . 2C ) . In Fig . 2B we consider a range in hydrophobic mismatch which may be realized , for instance , by varying the tail lengths in phosphatidylcholine ( PC ) lipid bilayers from PC10 to PC24 [82] , while in Fig . 2C we consider values of membrane tension up to the approximate rupture tension of lipid bilayers [48] , [82] . Fig . 2B shows that membrane-thickness-mediated interactions between chemoreceptor trimers vanish when the bilayer hydrophobic thickness matches the chemoreceptor hydrophobic thickness , and increase in magnitude with increasing magnitude of hydrophobic mismatch . Fig . 2C predicts that , for a hydrophobic mismatch corresponding to chemoreceptors and the cytoplasmic membrane of E . coli , membrane-mediated interactions between chemoreceptor trimers become more pronounced with increasing membrane tension , yielding an increased propensity for chemoreceptor clustering . The basic qualitative features of the interaction potentials in Fig . 2 can be understood from the thickness deformation field due to a single membrane inclusion . Consider , for simplicity , a cylindrical membrane inclusion with a hydrophobic thickness that exceeds the unperturbed bilayer hydrophobic thickness . The resulting thickness deformation decays approximately exponentially around the membrane inclusion with a characteristic decay length nm [33] , [48] . The decaying thickness deformation will overshoot [47] , [52] , leading to a zone of compression of the lipid bilayer , before the deformation eventually approaches zero ( S3 Figure ) . The attractive regime of membrane-mediated interactions in Fig . 2 corresponds to edge-to-edge separations of up to approximately , for which thickness deformations mainly overlap in the region of initial exponential decay and the overall deformation footprint of the two trimers is reduced compared to noninteracting trimers ( Fig . 1A ) . For edge-to-edge separations from approximately to , the compressed and expanded membrane regions induced by the two trimers strongly overlap , which results in frustration of membrane deformations and the repulsive regime in Fig . 2 . Finally , the noninteracting regime in Fig . 2 corresponds to edge-to-edge separations greater than approximately , for which there is only little overlap in the thickness deformations induced by the two trimers . The scale of the maximum interaction energies in Fig . 2 is set by the single-cylinder thickness deformation energy [33] for a radius R = 3 . 1 nm . Also , since , the strength of the attractive and repulsive regimes increases with the magnitude of the hydrophobic mismatch as in Fig . 2B . Moreover , the single-inclusion thickness deformation energy increases with membrane tension if , as is the case for chemoreceptors in the cytoplasmic membrane , the hydrophobic mismatch takes a positive value [48] , yielding an increase in the strength of membrane-mediated interactions with increasing membrane tension as in Fig . 2C . Fig . 2 shows that membrane-mediated interactions between chemoreceptor trimers are strongly directional , and reflect the three-fold symmetry of trimers . We find two dominant trimer configurations as a function of trimer separation: ( 1 ) In Fig . 2A , for trimer separations greater than nm , the tip-on configuration ( red inset ) is energetically most favorable . ( 2 ) For small trimer separations , smaller than nm , the face-on configuration ( blue inset ) is most favorable . These two regimes occur because the tip-on configuration yields the smallest edge-to-edge separation ( and thus the longest-range interactions , Fig . 1B left panel ) , while the face-on configuration maximizes the membrane area over which trimer-induced thickness deformations can overlap ( and thus provides the maximum interaction strength overall , Fig . 1B right panel ) . We estimate that the energy difference between tip-on and face-on configurations can be more than 10 for the minimum trimer separations allowed by steric constraints in the two configurations . In particular , membrane-mediated interactions favor the face-on trimer configuration at the observed separation nm ( grey vertical line ) as measured by electron cryo-tomography of chemoreceptor lattices in E . coli as well as other organisms [19] , [22] , in the presence of CheA and CheW . The face-on configuration of trimers predicted by our model for small trimer separations has been observed in chemoreceptor lattices in a variety of different organisms [19] , [20] and allows the formation of chemoreceptor-CheW-CheA complexes , yielding a well-defined trimer separation due to direct protein-protein interactions . Fig . 2 implies a scenario for the assembly of chemoreceptor lattices in which the tip-on trimer configuration is a gateway state yielding attraction between chemoreceptor trimers over several nanometers , with the directionality of membrane-mediated interactions ensuring that , at small separations , trimers are arranged in the face-on orientation allowing further stabilization through direct protein interactions mediated by CheA and CheW [19] , [20] . In particular , the interaction potentials in Fig . 2 suggest that the face-on trimer configuration found in chemoreceptor lattices [19] , [20] could be achieved through the sequence of gateway states shown in Fig . 3 . For large , the tip-on configuration is strongly favored ( for ease of visualization , the tip-on configuration is set as the zero of in Fig . 3 ) . As the trimer separation shrinks below the steric constraint on the tip-on configuration , the membrane deformation energy can be lowered further by a symmetric rotation of the chemoreceptor trimers ( S1 Video ) , ultimately yielding the observed face-on trimer configuration [19] , [20] as the lowest-energy configuration , thus ensuring correct assembly of chemoreceptor lattices . Consistent with the results in Fig . 2 , we find that the membrane-mediated interactions stabilizing the sequence of gateway states in Fig . 3 vanish for lipid bilayers matching the chemoreceptor hydrophobic thickness and increase with the magnitude of the hydrophobic mismatch ( Fig . 3A ) . Similarly , our model predicts that the reduction in membrane deformation energy associated with the sequence of gateway states in Fig . 3 increases with increasing membrane tension ( Fig . 3B ) . A simple arrangement of trimers in chemoreceptor lattices would be a close-packed hexagonal lattice structure ( Fig . 4 grey insets , S4A Figure ) in which each trimer has six nearest neighbors and , hence , the number of nearest-neighbor interactions is maximized . However , electron cryo-tomography has shown [19] , [20] that chemoreceptor trimers are not closely packed in chemoreceptor lattices but rather form a honeycomb lattice in which each trimer has three nearest-neighbors arranged in the face-on orientation ( Fig . 4 blue insets , S4B Figure ) , which allows formation of an extended lattice composed of chemoreceptor trimers , CheA , and CheW . To elucidate the stability of the observed face-on honeycomb-lattice architecture we calculated the energy per chemoreceptor trimer resulting from membrane-mediated interactions due to thickness deformations , , in face-on honeycomb , tip-on honeycomb ( Fig . 4 red insets , S4C Figure ) , and hexagonal lattices . We find that , while tip-on honeycomb and hexagonal lattices can be energetically favorable for large lattice spacings , both these structures are unstable to the formation of a face-on honeycomb lattice with small lattice spacing , which provides the minimum-energy lattice architecture ( Fig . 4A ) . This conclusion is robust with respect to variations in hydrophobic mismatch ( Fig . 4B ) and membrane tension ( Fig . 4B inset ) . In contrast , cylindrical membrane inclusions , which do not exhibit directional interactions , would yield the hexagonal lattice as the minimum-energy structure . Thus , the directionality of membrane-mediated interactions stabilizes the observed face-on honeycomb lattice architecture against the tip-on honeycomb and hexagonal lattice structures . Specifically , the three-fold symmetry of trimers allows honeycomb ordering of chemoreceptor lattices , and thus further stabilization of a well-defined lattice constant through direct protein interactions with CheA and CheW [19] , [20] . Fig . 4B predicts that , for the lattice spacings indicated by arrows in Fig . 4A , the strength of favorable interactions between chemoreceptor trimers in face-on honeycomb , tip-on honeycomb , and hexagonal lattices grows monotonically with increasing hydrophobic mismatch between lipid bilayer and chemoreceptors , as well as with increasing membrane tension . For the lattice spacings in Fig . 4A yielding a crossover from favorable ( ) to unfavorable ( ) lattice energies we obtain a more complex dependence of the lattice energy on bilayer hydrophobic thickness and membrane tension ( S5 Figure ) . In particular , for such crossover lattice spacings our model predicts favorable lattice energies for bilayer hydrophobic thicknesses exceeding the chemoreceptor hydrophobic thickness , with unfavorable lattice energies for bilayer hydrophobic thicknesses smaller than the chemoreceptor hydrophobic thickness . This can be understood by noting that the decay length increases with , thus shifting membrane-mediated interactions into the attractive regime if increases beyond , and vice versa . Finally , we note that the lattice energy due to membrane-mediated interactions between chemoreceptor trimers is dominated by nearest-neighbor interactions , with longer-range interactions only yielding minor shifts in the lattice energy ( S6 Figure ) . Our calculations imply that close-packed hexagonal lattices of chemoreceptor trimers are metastable in the sense that the hexagonal lattice structure is only a local minimum of the membrane deformation energy , with the global minimum provided by the face-on honeycomb lattice ( Fig . 4 ) . However , the membrane area per trimer in honeycomb lattices is greater than the membrane area per trimer in hexagonal lattices—by 50% if both lattice structures have the same trimer separation and by 15% for the trimer separations indicated by arrows in Fig . 4 . Thus , in situations where the clustering of chemoreceptor trimers is strongly constrained by the available membrane area , membrane-mediated interactions may yield hexagonal chemoreceptor lattices . On the basis of electron microscopy it has indeed been observed [83]–[85] that overexpression of chemoreceptors results in hexagonal lattices of trimers in the cytoplasmic membrane . The observed two-dimensional hexagonal lattices were distinct from the “zippered” cluster structures [83] also found in overexpression experiments , which strongly bend the membrane and form interdigitated protein contacts . In agreement with our model , in the case of overexpression the clustering of trimers , the stability of the lattice , and the two-dimensional hexagonal lattice architecture did not rely on the presence of CheA and CheW , although the presence of CheA and CheW yielded more ordered lattice structures and modified the lattice spacing [84] , [85] . The trimer orientation in the observed two-dimensional hexagonal lattices [84] , [85] is consistent with the hexagonal lattice architecture of trimers shown in Fig . 4 ( grey insets ) . As noted in the Models section , chemoreceptor trimers induce midplane deformations in addition to thickness deformations . While midplane interaction energies are typically negligible compared to thickness interaction energies ( see S1 Text section 2 ) , midplane deformations provide a simple mechanism for segregation of chemoreceptor trimers to the cell poles [86] . In particular , the energetic cost of trimer-induced curvature deformations depends on the interplay between the conical shape of chemoreceptor trimers [87] and the preferred curvature of the surrounding lipid bilayer: Since the average membrane radius of curvature at the poles of E . coli is approximately twice that of the midcell region , and both have the same sign as the radius of curvature of chemoreceptor trimers , midplane deformations may act as curvature sensors mediating localization of chemoreceptor trimers to the cell poles . The energy of trimer-induced midplane deformations can be estimated using a variety of different approaches [48] , [50] , [80] , [81] . Independent of the particular model formulation , we find that for a single chemoreceptor trimer the difference in midplane deformation energy between the poles and midcell of E . coli is well below ( see S1 Text section 2 ) . This suggests that curvature deformations are not able to localize individual chemoreceptor trimers to the cell poles . However , as described above , we also find that strong membrane-mediated interactions due to thickness deformations effectively bind chemoreceptor trimers into chemoreceptor lattices , which may be further stabilized by interactions with CheA and CheW . For a lattice composed of chemoreceptor trimers we estimate an energy difference ( 3 ) between the midcell and poles of E . coli in the regime of weak interactions due to midplane deformations , where the lower and upper bounds correspond to different model formulations ( see S1 Text section 2 ) . Thus , bilayer-trimer interactions yield only weak curvature sensitivity for small chemoreceptor lattices but can readily induce localization of large chemoreceptor lattices to convex regions of the cytoplasmic membrane such as the cell poles . Large chemoreceptor lattices composed of thousands of receptors ( for which Eq . ( 3 ) yields ) are indeed observed predominantly at the cell poles , while smaller chemoreceptor lattices are also found in the midcell regions [21]–[25] , [27]–[30] . Reconstitution of chemoreceptors in bilayer vesicles [36] and nanodiscs [35] has indicated that the signaling properties of chemoreceptors depend on the composition of lipid bilayers . Furthermore , modification of the transmembrane properties of chemoreceptors by site-directed mutagenesis has shown [37] that bilayer-chemoreceptor interactions influence chemotactic signaling . Within the simple “piston” model of chemotactic signaling such a coupling between chemoreceptor function and lipid bilayer properties arises naturally—specifically , the on and off states of chemoreceptors differ in their hydrophobic mismatch with the lipid bilayer and thus in their bilayer deformation energies . Assuming a uniform 0 . 16 nm difference in hydrophobic thickness between on and off states , our model predicts that for E . coli the membrane contribution to the total free-energy difference between on and off states of a single chemoreceptor trimer is greater than in magnitude , and can vary over more than with bilayer or trimer hydrophobic thickness ( S7 Figure ) . In agreement with experiments [35]–[37] we therefore find that shifts in the membrane contribution to the trimer transition energy due to modification of lipid composition or chemoreceptor transmembrane properties can dominate over shifts in the transition energy due to chemoreceptor methylation , which are of the order of per methyl group [88] . In addition , our model predicts a dependence of chemotactic signaling on membrane elastic properties such as membrane tension ( S7 Figure inset ) . In particular , we find that variation in membrane tension can shift the membrane contribution to the on-off transition energy by up to which , again , is comparable to shifts in the trimer transition energy due to chemoreceptor methylation [88] . We speculate that membrane-mediated interactions between chemoreceptor trimers could contribute to cooperativity among chemoreceptor trimers , complementing the contribution of direct protein interactions mediated by CheA and CheW [9] , [10] , [19] , [20]: Consider a chemoreceptor trimer in the on state , with a neighboring trimer in the off state ( Fig . 5A upper panel ) . Assuming that the two trimers induce distinct thickness deformations due to their different signaling states , membrane-mediated interactions are energetically unfavorable at small trimer separations [33] . If , however , both trimers are in the off state ( Fig . 5A lower panel ) , membrane-mediated interactions are strongly favorable . Thus , the presence of a neighboring trimer in the off state lowers , via membrane-mediated interactions , the free energy of the off state ( and similarly a neighbor in the on state lowers the free energy of the on state ) , potentially yielding membrane-mediated cooperativity among chemoreceptor trimers . In order to quantify the above mechanism for membrane-mediated cooperativity we calculated the membrane contribution to the free-energy difference between the on and off state of a chemoreceptor trimer , , for the trimer orientation [19] , [20] and separation [19] , [22] observed in chemoreceptor lattices ( Fig . 5B ) . Consistent with our results for a single trimer ( S7 Figure ) we find that there is a substantial membrane contribution to the transition energy of trimers in chemoreceptor lattices . Since chemoreceptors are functionally required to operate near zero transition energy [89] , this membrane contribution must be compensated by internal protein contributions to the transition energy . However , Fig . 5B also shows that membrane-mediated interactions can lower the transition energy by up to approximately depending on the activity state of neighboring trimers , or by approximately for each nearest-neighbor trimer in the off state . This cooperative shift in the transition energy is comparable to the shift in the trimer transition energy obtained by methylation of all 24 modification sites on a trimer [88] , and may therefore be relevant for the cooperative signaling properties of chemoreceptor lattices . Furthermore , Fig . 5B shows that the strength of the predicted cooperative interactions among chemoreceptor trimers is robust with respect to variations in hydrophobic mismatch ( Fig . 5B main panel ) and membrane tension ( Fig . 5B inset ) .
Fluorescence experiments have suggested a stochastic model for chemoreceptor lattice formation [28] , [30] in which self-assembly of chemoreceptor lattices proceeds by nucleation and growth , without direct cytoskeletal involvement or active transport . Lattices consist of trimers-of-dimers of chemoreceptors and require the linker/kinase CheA and the linker CheW [19] , [20] for their function . However , clustering of chemoreceptor trimers requires neither CheA nor CheW [27] , [28] . In common with other membrane proteins [31] , [32] , chemoreceptor trimers are expected to deform the surrounding lipid bilayer , leading to membrane-mediated interactions [33] , [34] between neighboring trimers . To quantify the role of membrane-mediated interactions in the assembly and architecture of chemoreceptor lattices we have developed a biophysical model of bilayer-chemoreceptor interactions . Our biophysical model of bilayer-chemoreceptor interactions shows that membrane-mediated interactions yield attractive interactions between chemoreceptor trimers over several nanometers and hence provide a biophysical mechanism for cluster self-assembly . Our model predicts that the tip-on orientation of a pair of chemoreceptor trimers is a “gateway” state during assembly , whereas at smaller trimer separations , membrane-mediated interactions favor the face-on orientation of each trimer pair also observed in the presence of CheA and CheW [19] , [20] . Furthermore , we predict that membrane-mediated interactions are strong enough to induce cluster formation even for a trimer concentration of only trimers per E . coli cell . This suggests a scenario for self-assembly in which membrane-mediated interactions produce clusters of chemoreceptors , which are further stabilized and ordered through protein interactions mediated by CheA and CheW . Since the range of membrane-mediated interactions is set by the elastic decay length of thickness deformations , which is a bilayer property , these conclusions do not rely on the detailed size and shape of trimers in our model of chemoreceptor trimers in Fig . 1B . In particular , for a given bilayer membrane the range of membrane-mediated interactions between trimers , measured in terms of the center-to-center distance between trimers , is determined by the edge-to-edge separation of trimers for each trimer configuration , yielding a longer ( shorter ) range of membrane-mediated interactions for larger ( smaller ) trimer sizes . In agreement with experimental observations [30] , the strongly favorable interactions between trimers at small separations in Figs . 2 and 3 are expected to yield an approximately exponential size distribution of chemoreceptor clusters [30] , [90] . In vivo electron cryo-tomography has revealed [19]–[25] that chemoreceptor lattices are not close-packed hexagonal arrays . Instead , chemoreceptor trimers form honeycomb lattices with a trimer at each vertex ( S4B Figure ) , and a well-defined face-on orientation of trimers [19] , [20] . Our model predicts that membrane-mediated interactions favor this face-on , honeycomb architecture of the lattice . In particular , we find that the three-fold symmetry and directionality of membrane-mediated interactions favor a honeycomb lattice ( three neighbors per trimer ) over a close-packed hexagonal lattice ( six neighbors per trimer ) . Thus , while interactions with CheA and CheW are expected to determine the observed separation of trimers in chemoreceptor lattices [19] , [22] and are likely to be adequate to define the observed lattice symmetry [73] , [91] , we find that membrane-mediated interactions can drive the formation of diffuse , less ordered chemoreceptor clusters [27] , [28] and further stabilize the face-on honeycomb architecture of chemoreceptor lattices involving CheA and CheW . These results rely only on generic properties of chemoreceptor trimers and the cytoplasmic membrane , specifically the three-fold symmetry of trimers and a hydrophobic mismatch between trimers and the cytoplasmic membrane . This generality suggests that membrane-mediated interactions may facilitate the consistently observed honeycomb architecture of chemoreceptor lattices [22] . Membrane-mediated interactions extend over a longer range than direct protein-protein interactions , but may be weaker in magnitude . Thus , membrane-mediated interactions in chemoreceptor lattices complement direct protein-protein interactions , yielding robustness of the overall chemoreceptor lattice architecture against local disruption . Indeed , it has been observed [9] , [26] , [27] that chemoreceptor lattices can exhibit variable stoichiometries of chemoreceptors , CheA , and CheW . Our model predicts that membrane-mediated interactions can help to establish the proper orientation of neighboring trimers and the overall honeycomb lattice symmetry even at suboptimal protein stoichiometries , and thereby help to preserve lattice symmetry and stability . Conversely , it has been found [83]–[85] that overexpression of chemoreceptors can yield a two-dimensional hexagonal rather than a honeycomb lattice of trimers . In agreement with these observations , our model reveals that the honeycomb lattice structure is favored by the directionality of membrane-mediated interactions at moderate trimer densities while the hexagonal lattice structure is favored at high chemoreceptor densities . Our model of chemoreceptor trimers in Fig . 1 assumes that chemoreceptor trimers induce bilayer deformations and possess a three-fold symmetry . The former assumption is thought [31] , [32] to be a generic feature of transmembrane proteins such as chemoreceptors . The latter assumption is only justified if the three-fold symmetry of trimers , observed most directly in the cytoplasmic region of trimers , is also present in the transmembrane region of trimers . Electron cryo-tomography of chemoreceptor trimers has suggested [19] , [20] , [73] that the chemoreceptor dimers forming a trimer spread apart within the membrane . This may allow penetration of lipids into chemoreceptor trimers and , hence , membrane-mediated interactions within trimers . We did not consider such interactions here . Instead , we focused on membrane-mediated interactions between trimers which , within our model , might either correspond to compact chemoreceptor complexes or , alternatively , to lipid-chemoreceptor complexes . We note , however , that the penetration of lipids into chemoreceptor trimers may facilitate fluctuations in the relative positions of dimers within trimers , thereby reducing the rigidity of trimer shape . Such fluctuations could have interesting effects . For instance , while fluctuations in trimer shape are expected to reduce the directionality of membrane-mediated interactions , they could also increase the strength of membrane-mediated interactions by allowing a more favorable interface between neighboring trimers . Similarly , fluctuations in the structure of chemoreceptor trimers in the cyto- or periplasmic trimer regions could give rise to direct trimer-trimer interactions , which would compete with membrane-mediated interactions between trimers . Our model of bilayer-chemoreceptor interactions suggests that localization of large chemoreceptor lattices to the cell poles is simply a consequence of the conical shape of individual chemoreceptor trimers [87] , and neither requires interactions with CheA and CheW [27] nor curvature-mediated interactions among trimers . In agreement with experimental observations [21]–[25] , [27]–[30] , our model implies that large chemoreceptor clusters will tend to localize at the cell poles , while smaller chemoreceptor clusters can be distributed throughout the midcell regions . This mechanism for localization of large chemoreceptor lattices due to curvature sensing by individual chemoreceptor trimers is to be contrasted with a previously proposed mechanism [86] which assumes that trimers interact to yield a non-zero global intrinsic curvature of chemoreceptor lattices . A distinguishing difference between the localization mechanism proposed here and in Ref . [86] is that , according to the latter , chemoreceptor clusters should have a finite characteristic size set by the energy balance between short-range attraction and curvature-mediated long-range repulsion between trimers , whereas our model indicates that curvature-mediated interactions are too weak to limit cluster size in the absence of CheA and CheW . Fluorescence experiments [28] , [30] measuring chemoreceptor cluster size in the absence of CheA and CheW may be able to distinguish between these two related scenarios for curvature-driven localization of large chemoreceptor lattices . FRET experiments have revealed [8] , [9] that chemoreceptors signal in cooperative teams of coupled trimers [9] , [11]–[16] . Cooperative interactions among neighboring trimers are believed to be mediated by CheA and CheW [9] , [10] , [19] , [20] . Our model of chemotactic signaling shows that , provided there is a substantial change in chemoreceptor hydrophobic thickness upon signaling , membrane-mediated interactions between chemoreceptor trimers [85] can in principle yield cooperative interaction energies of the order of several . This would be sufficient [92] to account at least in part for the observed cooperative signaling properties of chemoreceptor lattices . Indeed , electron cryo-tomography indicates that honeycomb lattices of chemoreceptor trimers are somewhat disordered , with the degree of disorder being a matter of debate [21]–[25] . While interactions between chemoreceptor trimers mediated by CheA and CheW [9] , [10] , [19] , [20] rely on a regular lattice structure , membrane-mediated interactions are less sensitive to defects in the chemoreceptor lattice . Thus , membrane-mediated interactions may increase the robustness of cooperative signaling teams , and complement cooperative interactions mediated by CheA and CheW [9] , [10] , [19] , [20] . Consistent with our biophysical model of chemotactic signaling it has been found using homo-FRET [87] , [93] that the in vivo signaling response of chemoreceptors depends on membrane-mechanical properties such as membrane tension . However , homo-FRET has so far not produced any evidence for cooperativity among chemoreceptor trimers in the absence of CheA and CheW [94] . Chemoreceptor clusters formed in the absence of CheA and CheW are more diffuse than chemoreceptor lattices formed in the presence of CheA and CheW [27] , [28] , which may substantially reduce membrane-mediated cooperativity . Our model of chemotactic signaling predicts that shifts in the membrane contribution to the total free-energy difference between on and off states of chemoreceptor trimers due to changes in membrane composition or membrane tension can be comparable to shifts in the chemoreceptor transition energy due to receptor methylation [88] , and can therefore be functionally relevant . In agreement with these predictions , it has been found that modifying the composition of lipid bilayers [35] , [36] or bilayer-chemoreceptor interface [37] affects chemotactic signaling . In particular , changes in lipid composition can strongly bias chemoreceptors towards the active or inactive state [35] , [36] , and the baseline signaling state of chemoreceptors can be controlled by site-directed mutagenesis of chemoreceptor transmembrane helices [37] . Thus , in analogy to gramicidin [47] , [71] and mechanosensitive [49] , [72] channels , systematic variation of the membrane lipid composition , the chemoreceptor hydrophobic thickness , or membrane-mechanical properties such as membrane tension may allow quantitative experimental tests of our biophysical model of the role of membrane-mediated interactions in the assembly and architecture of chemoreceptor lattices , as well as our speculation of a membrane-mediated contribution to chemotactic signaling and cooperativity .
|
The chemotaxis system allows bacteria to respond to minute changes in chemical concentration , and serves as a paradigm for biological signal processing and the self-assembly of large protein lattices in living cells . The sensitivity of the chemotaxis system relies crucially on cooperative interactions among chemoreceptor trimers , which are organized into intricate honeycomb lattices . Chemoreceptors are membrane proteins and , hence , are expected to deform the surrounding lipid bilayer , leading to membrane-mediated interactions between chemoreceptor trimers . Using a biophysical model of bilayer-chemoreceptor interactions we show that the membrane-mediated interactions induced by chemoreceptor trimers provide a mechanism for the observed self-assembly of chemoreceptor lattices . We find that the directionality of membrane-mediated interactions between trimers complements protein-protein interactions in the stabilization of the observed honeycomb architecture of chemoreceptor lattices . Our results suggest that the symmetry of membrane protein complexes such as chemoreceptor trimers is reflected in the anisotropy of membrane-mediated interactions , yielding a general mechanism for the self-assembly of ordered protein lattices in cell membranes .
|
[
"Abstract",
"Introduction",
"Models",
"Results",
"Discussion"
] |
[
"physics",
"cell",
"signaling",
"biochemistry",
"cell",
"motility",
"signal",
"transduction",
"cell",
"biology",
"membrane",
"receptor",
"signaling",
"proteins",
"biophysics",
"theory",
"transmembrane",
"receptors",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"computational",
"biology",
"molecular",
"cell",
"biology",
"biophysics",
"chemotaxis"
] |
2014
|
The Role of Membrane-Mediated Interactions in the Assembly and Architecture of Chemoreceptor Lattices
|
The protozoan parasite Leishmania donovani ( LD ) reduces cellular cholesterol of the host possibly for its own benefit . Cholesterol is mostly present in the specialized compartment of the plasma membrane . The relation between mobility of membrane proteins and cholesterol depletion from membrane continues to be an important issue . The notion that leishmania infection alters the mobility of membrane proteins stems from our previous study where we showed that the distance between subunits of IFNγ receptor ( R1 and R2 ) on the cell surface of LD infected cell is increased , but is restored to normal by liposomal cholesterol treatment . We determined the lateral mobility of a membrane protein in normal , LD infected and liposome treated LD infected cells using GFP-tagged PLCδ1 as a probe . The mobility of PLCδ1 was computationally analyzed from the time lapse experiment using boundary distance plot and radial profile movement . Our results showed that the lateral mobility of the membrane protein , which is increased in infection , is restored to normal upon liposomal cholesterol treatment . The results of FRAP experiment lent further credence to the above notion . The membrane proteins are intimately linked with cellular actin and alteration of cellular actin may influence lateral mobility . We found that F-actin is decreased in infection but is restored to normal upon liposomal cholesterol treatment as evident from phalloidin staining and also from biochemical analysis by immunoblotting . To our knowledge this is the first direct demonstration that LD parasites during their intracellular life cycle increases lateral mobility of membrane proteins and decreases F-actin level in infected macrophages . Such defects may contribute to ineffective intracellular signaling and other cellular functions .
The protozoa parasite , Leishmania donovani , ( LD ) replicates within the macrophage of the mammalian host [1] . The parasite during their intracellular life cycle causes wide variety of defects in cellular physiology like decrease in membrane cholesterol [2] with concomitant increase in membrane fluidity [3] . LD infectedmacrophage are unable to stimulate antigen specific T cells [3] and display defective IFNγ receptor 1 ( R1 ) and receptor 2 ( R2 ) subunit assembly [4] . Kala-azar patient showed progressive decrease in serum cholesterol as a function of splenic parasite load [5] . Interestingly , defective T cell stimulating ability and IFNγR assembly of infected macrophage can be corrected by liposomal cholesterol treatment [2] , [4] . Cholesterol is one of the main constituent of the cell membrane and is important for raft assembly [6] . There are controversial reports on cholesterol depletion and membrane protein mobility , in some cases cholesterol depletion suppresses membrane protein mobility [7]–[9] while in others it increases the lateral mobility of the membrane protein [10] , [11] . Several studies showed that decreased protein mobility on cholesterol depletion is due to the changes in the architecture of the underlying cytoskeleton [12] , [13] . The formation of an immunological synapse between T cells and antigen presenting cells ( APC ) is recognized as a key event for activation of T cell [14]andactin cytoskeleton plays an important role in T cell activation [15] . Cholesterol , an important constituent oflipid raft [6] , is intimately involved in the dynamics of immune synapse formation [16] . Previously we showed that splenic macrophages of infected hamster are incapable to form immunological synapse while splenic macrophages of liposomal cholesterol treated infected hamsters can form immunological synapse [2] . The duration of contact between T cell and APC is also important for T cell activation . The mature synapse lasts for several hours and is thought to be important for sustained signalling [17] . There is a report that slow moving peptide-MHC complex but not fast moving ones can form immunological synapse [18] . It has been shown in Plasmodium falciparum infection that the lateral mobility of erythrocytes membrane protein is related to the infective stages of the parasite [19] . 1-phosphatidylinositol 4 , 5-bisphosphate phosphodiestares delta 1 ( PLCδ1 ) is a membrane protein that catalyzes hydrolysis of phosphatidylinositol 4 , 5 biphosphate to generate diacylglycerol and inositol 1 , 4 , 5 triphosphate ( IP3 ) . Using PLCδ1 as a representative of membrane proteins , we show that under parasitized condition the lateral mobility of PLCδ1 is significantly increased coupled with reduced actin polymerization . The increased lateral mobility of PLCδ1 observed in LD infected macrophages is restored to normal upon liposomal cholesterol treatment . The enhance dynamic of membrane proteins may contribute to the defective signal transduction leading to defective cellular function .
Use of mice was approved by the Institutional Animal Ethics Committee of Indian Institute of Chemical Biology , India . All animal experimentations were performed according to the National Regulatory Guidelines issued by CPSEA ( Committee for the Purpose of Supervision of Experiments on Animals ) , Ministry of Environment and Forest , Govt . of India . The protocol number is SDR/SYR/2007 . LipoFECTAMINE , alexa 488 phalloidin and FCS ( Fetal calf serum ) were purchased from Invitrogen . Luria-Bertani media purchased from Merck , India . DNA preparation minikit was purchased from Qiagen . Penicillin-streptomycin , kanamycin , chloramphenicol , sodium bicarbonate , β mercapto-ethanol , RPMI-1640 , M199 , Hoechst 33258were purchased from Sigma Aldrich ( St . Louis , MO ) . Cholesterol , phosphatidyl choline and cholesterol analogue were purchased from Avanti Polar Lipids . β-Actin antibody ( mouse monoclonal IgG1 ) and secondary antibody ( goat anti-mouse-HRP ) were obtained from Santa Cruz and Bangalore Genei , Bangalore , India respectively . plcδ1-gfp plasmid is a kind gift of Dr . TamasBalla , NIH , USA . RAW 264 . 7 ( murine macrophage cell line ) was used for in vitro experiments . For convenience RAW 264 . 7 was defined as macrophages ( MΦ ) . The cell line was maintained in RPMI-1640 medium supplemented with 10% FCS and β-mercaptoethanol ( 5×10−5 M ) at 37°C with 5% CO2 in a humidified atmosphere . Leishmania donovani strain AG83 ( MHOM/IN/1983/AG83 ) , originally obtained from Indian kala-azar patients , was maintained in Golden Hamsters as described previously [20] . Promastigotes obtained after transforming amastigotes from spleen of infected animals were maintained in culture M199 supplemented with 10% FCS at 22°C . The culture was replenished with fresh medium every 72–96 h . RAW 264 . 7cells ( 104 cells/well ) were platedinto 8 chambered coverslips ( BD Bioscience ) for 14 h . The non-adherent cells were removed by washing . The cells were transfected with plcδ1-gfp plasmid using LipoFECTAMINE following the Invitrogen LipoFECTAMINE kit protocol . After 6 h , the cells were washed and complete medium was added . RAW 264 . 7 cells were infected as described previously [4] . Briefly RAW 264 . 7 cells ( 105/106 ) were allowed to adhere coverslips/petri dish for 24 h at 37°C under 5% CO2 atmosphere , after which the non-adherent cells were removed by gentle washing with serum-free medium . The adherent cells , after overnight incubation in complete medium , were challenged with stationary phase LD promastigotes at a cell to parasite ratio of 1∶10 and incubated for 6 h at 37°C . Excess parasites were then washed off with serum-free medium . The cells were then incubated further for 48 h . At end points the cover slips were washed with PBS , dried , fixed with 100% methanol , and stained with 10% Giemsa . The intracellular parasites were enumerated microscopically and the results were expressed as % infected RAW 264 . 7 cells as well as the number of parasites/100 RAW 264 . 7 cells . For convenience , LD infected RAW264 . 7 cells were defined as I-MΦ . Liposomal cholesterol and liposomal cholesterol analogue were prepared using cholesterol/cholesterol analogue and phosphatidylcholine ( PC ) at a molar ratio of 1 . 5∶1 as previously described [2] . Briefly , 5 . 8 mg cholesterol/cholesterol analogue ( 4-cholestene-3-one ) and 8 mg PC in chloroform were mixed and a thin film was prepared; subsequently , the film was dissolved in 1 ml saline and sonicated ( Microson Ultrasonic cell disruptor with a Misonix 2-mm probe ) at 4°C three times for 1 min each time at maximum output . The LD infected RAW264 . 7 cells ( 104 cell/200 µl ) were incubated with 10 µl liposomes for 20 h at 37°C . The cells were then washed three times in serum-free RPMI 1640 medium and finally resuspended in 10% FCS containing RPMI 1640 . For convenience , liposomal cholesterol treated LD infected cells as I-MΦ-CL , and liposomal cholesterol analogue treated LD infected cells as I-MΦ-AL . The plcδ1-gfp transfected RAW 264 . 7 cells were visualized under Confocal microscope . The cells were incubated with 1 µg/ml Hoechst 33258 solution for 2 minutes and washed off with PBS ( Phosphate buffered saline , pH 7 . 2 ) to visualise the nucleus . The cells were then kept in complete medium ( RPMI 1640 containing 10% FCS ) and live cell images of the cells were taken for 1 minute with 8 seconds interval . GFP was imaged using a wide field microscope ( model: Andor Spinning Disc Confocal Microscope , Olympus ) . The fluorochrome was excited with a mercury arc lamp . For alexa 488 excitation and emission wavelength was488 and 525 nm . For Hoechst 33258 these were 405 and 447 nm respectively . Images , obtained using identical exposure times for cells subjected to various treatments in each experiment , were collected using a 60X1 . 42 NA Plan-ApoN objective and captured using Andor IXON897EM CCD camera . The spatial distribution of the expression of PLCδ1-GFP in normal , infected and liposomal cholesterol treated infected cells was computationally analyzed using boundary distance analysis as described previously [21] . Briefly , the differences in the shape and size of the cells were adjusted by normalizing the spatial co-ordinates of the cell . The normalization were done on a one dimensional scale with the following features: ( i ) the ‘center’ of the cell has a radial distance of 0 ( ii ) points on the boundary of a cell to be a radial distance of 1 ( iii ) points in the interior of a cell have a radial distance between 0 and 1 , depending on how close they are to the boundary of the cell ( iv ) points outside the boundary of the cell have a radial distance of more than 1 , with further points having greater distance . The co-ordinate system was computed separately for each cell . Detecting the motion of the PLCδ1-GFP at a given location in a cell is a hard problem because we cannot track individual particles over time . From live cell imaging , we obtained a time sequence of images , in a short space of time ( 1 minute ) , each of which the expression distribution has changed slightly from the previous one . From this sequence , we measured the temporal flux , in GFP expression , as follows ( 1 . 1 ) Where g ( x , y , t ) is the expression of the PLCδ1-GFP at location ( x , y ) and timepoint t . The movement ( flux ) measured , M ( x , y ) depends on how much the expression level changes from one timepoint to the next . In order to resolve the spatial distribution of movement , we have used the radial mapping concept described in the previous section for ( 1 . 1 ) , but this time mapping the movement measure M ( x , y ) instead of the expression levels . For increased comparability , we again normalized the movement measure by the overall level of expression for each cell . The highest level of expression or peak expression of GFP occurs at the membrane for each cell . Thus we calculated P , peak level of expression of GFP for each cell . To compare these measurements across groups , we fitted a model of the form ( 1 . 2 ) Where Pij is the peak expressions of GFP for cell j in study i . The model terms include μ , the baseline mean for N-MΦ and si , an effect due to imaging study i . Since the factors that can influence the imaging study , such as ambient lighting , level of plcδ1-gfp transfection etc . , are unmeasured , we assume that this effect is random and has a zero mean Gaussian distribution . The main quantity of interest is tij , the effect of infection or treatment and finally εij is measurement error , assumed to have an independent mean zero Gaussian distribution . The Parameter estimates and hypothesis tests were carried out by fitting the corresponding mixed effects model using the R computing platform ( www . r-project . org ) . FRAP measurement was done as described [22] . Briefly , GFP was imaged using a wide field microscope ( model Andor Spinning Disc Confocal Microscope , Olympus ) . The fluorochrome was excited with a mercury arc lamp , using excitation wavelength of 488 nm; the emission wavelength was 525 nm . Cells were kept at 37°C in a humidified chamber . A 60X 1 . 42 NA Plan-ApoN objective was used with the confocal pinhole set at 1–2 Airy units . Photobleaching of GFP was performed with the 488-nm laser line at 30 mW power in a 32 . 85 µm2 rectangular region of interest . Pre- and post-bleach images were monitored at low laser intensity ( 14 mW ) . Fluorescence recoveries in the bleached region and overall photobleaching in the whole cell during the time series were quantified using the iQ2 . 7 software . In the confocal microscope , a prebleach series ( usually about 10 images ) at low illumination was acquired to measure the fluorescence equilibrium before photobleaching . One or more spots or regions of interest ( ROI ) were illuminated with high intensity to photobleach the area of interest . Finally , a post bleach image series were acquired to measure the recovery kinetics of fluorescent light intensity . The images were captured continuously ( real time ) for 30 s . Mobility of the PLCδ1-GFP were extracted from equation ( 1 . 3 ) . ( 1 . 3 ) where ROI is the Region of Interest , F ( t ) is the intensity of ROI at time t , F0 is the intensity of ROI immediately after bleach , Ff is the intensity of ROI at saturation , K is the recovery rate constant . The cells were permeabilized and fixed with 2% paraformaldehyde containing 0 . 1% Triton X 100 . To visualize actin , the cells were then stained with alexa 488 conjugated phalloidin for 30 min as described by manufacturer . After , washing the cells were mounted on a mounting medium containing DAPI . Alexa 488 fluorescence was imaged using a wide field microscope ( model Andor Spinning Disc Confocal Microscope , Olympus ) . The fluorochrome was excited with a mercury arc lamp . For alexa 488 excitation and emission wavelength was 488 nm and 525 nm respectively . Similarly for DAPI , excitation and emission wavelengths were 405 nm and 447 nm respectively . Images were collected using a 60X 1 . 42 NA Plan-ApoN objective and captured using Andor IXON897EM CCD camera . These were obtained using identical exposure times for cells subjected to various treatments in each experiment . The cellular F-actin ( polymerized form ) was isolated by modifying the method described previously [23] . Briefly , the cells were pelleted by centrifugation at 14000 rpm and resuspended in actin stabilisation-extraction buffer ( 0 . 1 M Pipes pH 6 . 9 , 30% glycerol , 5% DMSO , 1 mM MgSO4 , 10 µg/ml antiprotease cocktail , 1 mM EGTA , and 1% Triton X-100 ) at room temperature for 20 min and centrifuged at 14000 rpm for 15 min [24] . The pellet comprising the F-actin was dissolved in actin extraction buffer ( 2 mMTris-HCl pH-8 , 1 mM Na2 . ATP , 0 . 2 mM CaCl2 , 0 . 5 mM DTT ) [25] . For assaying total cellular actin , the cells were lysed with 1X RIPA Buffer from Cell signalling technology , Danvers , MA ( 20 mMTris-HCl pH 7 . 5 , 150 mMNaCl , 1 mM Na2EDTA , 1 mM EGTA , 1% NP-40 , 1% sodium deoxycholate , 2 . 5 mM sodium pyrophosphate , 1 mM beta glycerophosphate , 1 mM Na3VO4 , and 1 mg/ml leupeptin , with recommended addition of 1 mM PMSF immediately before use ) and centrifuged at 14000 rpm for 15 min [4] . The protein concentration was estimated by Bradford method using a Protein Estimation Kit ( Merck Genei , Mumbai , India ) . For immunoblotting , equal amounts of protein were loaded to SDS-PAGE ( 10% gel ) and electrotransferred to nitrocellulose membranes ( Millipore , Bangalore , India ) in a transfer buffer consisting of 20 mMTris-HCl , 150 mM glycine , and 20% methanol . Membranes were blocked overnight at 4°C in 5% BSA , probed with primary antibody against β-actin ( 1∶1000 dilution , mouse monoclonal IgG1 , Santa Cruz ) for 4 h at room temperature , and incubated with HRP-conjugated secondary antibody ( 1∶1500 dilution , Genei , Bangalore , India ) for 2 h at room temperature . The chemiluminescence signal was detected using Super Signal West Pico Chemiluminescent Substrate ( Pierce , Rockland , IL ) . The F-actin and total actin expressions were analyzed by image-J .
For this investigation RAW264 . 7 cells were used as host cell for infection with LD . The number of intracellular LD and integrity of the nucleus in infected MΦ ( I-MΦ ) were determined by staining with Hoechst 33342 or DAPI . The intracellular parasite number , verified by Giemsa staining ( S1 Figure ) , was ∼8–9/cell , similar to the results obtained by Hoechst 33342 or DAPI stain ( Fig . 1 ) . It was clear that the morphology of the cells and the integrity of nucleus did not change upon infection . Normal MΦs ( N-MΦ ) were transfected with plcδ1-gfp plasmid and the distribution of GFP was measured by confocal microscope . It was observed that green fluorescence was localized predominantly on the cell surface indicating cell surface expression of PLCδ1 ( S2A Figure ) . N-MΦs were transfected with plcδ1-gfp plasmid and then infected with LD . The expression of PLCδ1 in I-MΦwas found to be localized predominantly on the cell surface ( Fig . 1 ) . Similarly expression of PLCδ1 in liposomal cholesterol treated infected macrophages ( I-MΦ-CL ) was predominantly localized on the cell surface ( S2B Figure ) . The expression of PLCδ1 in N-MΦ , I-MΦ and I-MΦ-CL was quantified by computational method . The boundary distance co-ordinates were computed for each cell using the Euclidean distance transform after presmoothing and oversampling of cellular boundaries to normalize with respect to variations in cell shape and size . The general shape profile based on GFP expression in N-MΦ , I-MΦ and I-MΦ-CL were similar ( Fig . 2A ) . An adaptive piecewise linear model is used to compare expression gradients in intra , peri and extra cellular zones . We have considered the center of the cell as ‘0’ and the cell surface as ‘1’ on the x-axis . The distribution of PLCδ1 are relatively flat in the interior of the cell ( between 0 and 1 on the axis ) , followed by a sharp rise at the plasma membrane ( at 1 on the x-axis ) and finally a sharp drop off ( points >1 on the x-axis ) outside the cell ( Fig . 2B ) . The average expression curves show the same basic spatial distribution in N-MΦ , I-MΦ and I-MΦ-CL: low near the center of the cell , peak near the boundary and then a drop to ∼‘0’ some distance outside the boundary ( Fig . 3 ) . Differences in the highest level of PLCδ1expression or peak expression “P” across the cell was analyzed using the model in equation ( 1 . 2 ) . There was some variation in the location of the PLCδ1 expression peak across cells: this could be due to shape deformable cell movement observed during live cell imaging . The expressions of PLCδ1 on the cell surface of N-MΦ , I-MΦ and I-MΦ-CL were 248 . 97 , 241 . 76 and 236 . 04 a . u . respectively ( Table 1 ) . It appears that there was no significant difference in the expression of PLCδ1 in the above cell type . We studied the lateral mobility of PLCδ1 in LD infection using computational analysis . The radial profile of PLCδ1 movement showed that the movement is highest near the cell membrane irrespective of infection status of the cell . The highest movement at the cell boundary is denoted as peak movement or “P” . It was observed that peak movement is higher in the I-MΦ as compared to N-MΦ , whereas in I-MΦ-CL it was comparable to that in N-MΦ ( Fig . 4 ) . To confirm this result , we carried out an ANOVA analysis of PLCδ1 movement across cells , as described in equation ( 1 . 2 ) . The fitted model indicates that movement of PLCδ1 at the cell membrane is significantly higher in I-MΦ as compared to N-MΦ ( p-value = 0 . 04 ) and there was no significant difference between N-MΦ and I-MΦ-CL ( p-value = 0 . 72 ) ( Table 2 ) . The lateral mobility of PLCδ1 wasstudied in live cells at 37°C . The cells were bleached and fluorescence recovery of the bleached region was measured up to 30 s ( S3A Figure ) . It was observed that the extent of recovery was similar in N-MΦ , I-MΦ , I-MΦ-CL and I-MΦ-AL ( S4 Figure ) though the rate of recovery was different . The analysis of fluorescence recovery kinetics showed that the diffusion coefficient of PLCδ1 in N-MΦ , I-MΦ , I-MΦ-CL and I-MΦ-AL was 1 . 4±0 . 2 , 2 . 5±0 . 5 , 1 . 6±0 . 3 and 2 . 6±0 . 5 µm2 s−1 respectively ( Fig . 5 ) . The photobleach effect was specific because in the unrelated region , mean fluorescence intensity was constant . To show that the recovery was predominantly from the plasma membrane , cytosolic protein was also bleached as a control . It was observed that the recovery of PLCδ1 was very poor almost not detectable ( S3B Figure ) . We studied the actin cytoskeleton in N-MΦ , I-MΦ and I-MΦ-CL . The integrity of the cytoskeleton protein actin was measured by confocal microscopy after staining with fluorescence label phalloidin , which binds to filamentous actin ( F-actin ) . The actin filamentare clearly visible in N-MΦ ( Fig . 6A ) . However , in I-MΦactin filaments were reduced; insteadpossible actin depolymerisationwas noted ( Fig . 6B ) . Interestingly , I-MΦ-CLdid display actin filaments ( Fig . 6C ) . The computational analysis of phalloidin staining showed about 25% decrease in F-actin in I-MΦ and 9% increase in I-MΦ-CL as compared to N-MΦ . We next quantified F-actin by western blot and the result was expressed with respect to total actin . It was observed that the total actin level was similar in N-MΦ , I-MΦ and I-MΦ-CL ( S5B Figure ) . Though F-actin level was reduced in I-MΦ , it reappeared upon liposomal cholesterol treatment ( S5A Figure ) . There is about 30% decrease of F-actin in I-MΦ as compared to N-MΦ and in I-MΦ-CL F-actin was comparable ( S5C Figure ) .
The relation between cholesterol depletion from the membrane and lateral mobility of the membrane proteins remains a contentious issue . There are reports to suggest that the mobility of membrane proteins is decreased upon cholesterol depletion [7]–[9] , which may be due to formation of solid gel-like cluster in the membrane [26] . Again an increase in lateral mobility of membrane proteins like CD44 and wild type-H-Ras is reported after cholesterol depletion ( 10–12 ) . There is an interesting observation in neuronal cells where lateral mobility of nicotinic acetylcholine receptor was found to be governed by the receptor composition and local domain and cell type [27] . Here we studied the lateral mobility of PLCδ1 as prototype of membrane protein . PLCδ1 is a raft associated protein having cytoskeleton interacting ability [28] . It is a well characterized protein in terms of domain sequences and membrane interacting domains . UNIprot entry of PLCδ1 from Rattusnorvegicus ( ID: P10688 ) suggests existence of an N terminal pleckstrin homology ( PH ) domain ( region 21–130 ) followed by two EF-hand domains ( region 140–211 ) and two Phosphatidylinositol-specific phospholipase C domains ( region 296–609 ) [29] . The C-terminal end of PLCδ1 is also flanked by a C2 domain ( region 630–720 ) [30] . Both PH and C2 domains are known to involve in targeting proteins to cell membranes [31] . The transfection of plcδ1-gfp is well known to be specifically localized to plasma membrane with very little localization in the cytosol and nucleus [32] . The disease visceral leishmaniasis is characterized by immune suppression . Previously we showed that there is a defective synapse formation between T cells and parasitized MΦ [3] . The T-cell receptor recognizes Peptide-MHC complex in the context of antigen presenting cells ( APC ) like macrophages and dendritic cells [33] . The duration of contact between T-cells and APCs is critical for T-cell activation . It is an exciting proposition to study the lateral mobility of peptide-MHC complex to explain defective T-cell function in leishmaniasis . But there are issues that tend to discourage undertaking such studies . The infected macrophages show decreased affinity of MHC II towards peptide [34]; thus it will be difficult to interpret the defective T-cell stimulation due to increase in lateral mobility of MHC or lack of immunogenic peptides in association with MHC II protein . Furthermore there is a report on the existence of two species of peptide-MHC complex; the slower complex can form effective synapse but not the faster moving one [18] . It may be recalled that MHC class I when coupled with GFP ( GFP-tagged H-2Ld ) shows relatively low diffusion coefficient [35] . The diffusion coefficient may be influenced by the protein dimensions [36] . In this context PLCδ1 offered advantages as it is a single chain membrane protein as opposed to MHC protein which contains two chains . Our study showed that transfection of plcδ1 led to over expression of PLCδ1 in the membrane of RAW 264 . 7 cells , similar to that reported in CHO cells by others [37] . In the latter case , despite substantial increase in PLCδ1 in membrane , IP3 production was only marginally increased and it still needed a stimulus to generate IP3 [37]; therefore it is unlikely that the lipid composition in transfected cells may be altered in our case . Two complementary techniques have been exploited in this investigation to enhance our understanding in membrane protein dynamics under parasitized condition . It was observed that radial profile of PLCδ1 movement showed an increase in I-MΦ which was restored to normal upon liposomal cholesterol treatment ( Fig . 4 ) . This methodology has a limitation because we cannot track individual particles over time , as monitoring the motion of PLCδ1 at a given location in a cell is not easy [21] . FRAP study also showed an increase in the diffusion coefficient of PLCδ1 in I-MΦ as compared to N-MΦ ( Fig . 5 ) . The treatment of infected cells with liposomal cholesterol but not with liposomal cholesterol analogue ( 4-cholestene-3-one ) restored lateral mobility of PLCδ1 ( Fig . 5 ) . The amphiphilic properties of cholesterol are provided by the hydrophilic 3β-hydroxy group and the hydrophobic tetracyclic ring with the isooctyl side chain at C-17 . The reason we have included 4-cholestene-3-one as analogue of cholesterol is because it lacks the OH function of cholesterol which forms hydrogen bond with amide of sphingolipids , important for alignment of cholesterol in the membrane [38] . There was another study from our group where we showed that liposomal cholesterol but not the analogue of cholesterol favors raft assembly in infected macrophages [3] . The importance of 3β-hydroxyl function was further substantiated from another study by our group where we could show that the assembly of IFNγ receptor subunits ( R1 and R2 ) on the surface of leishmania infected macrophages could only be restored by liposomal cholesterol , not by cholesterol analogue or DPPC-liposomes [4] . Unlike the exquisite specificity of 3β-hydroxyl function of cholesterol , side chain modification appears still permissible because cholesterol may be replaced by desmosterol for packing in the membrane [39] . Several studies showed the effect of cholesterol on mechanical properties of a cell through the underlying cytoskeleton [12] , [13] . Changes in membrane cytoskeleton adhesion are expected to have a major impact on numerous cell functions [40] . The role of cytoskeleton meshwork as a barrier for lateral mobility of transmembrane proteins was established based on the findings that disorganized cytoskeleton meshwork favors faster lateral diffusion [41] . Kwik and coworkers [42] demonstrated that decrease in membrane protein mobility is associated with changes in architecture of the underlying actin network . F-actin plays an important role in restricting lateral mobility of membrane proteins in neuronal [43] and other cell types [44] , which indicates direct or indirect tethering of membrane proteins to the cytosketeton . Quantification of such interaction using optical technique , i . e . TIFR may be of great use as reported by Huhn and Pollard , but such method is restricted to purified actin [45] . Here we show the effect of cytoskeleton actin filament disruption on LD infection and its restoration by liposomal cholesterol treatment ( Fig . 6 ) . We quantified F-actin by western blot and the result is expressed with respect to total actin . It was observed that total actin is essentially similar in N-MΦ , I-MΦ and I-MΦ-CL; F-actin is reduced ( 30% ) only in I-MΦ but is restored upon liposomal cholesterol treatment ( S5A–B Figure ) . The comptutational analysis of phalloidin staining showed that there is about 25% decrease in F-actin in I-MΦ and which is restored to normal in I-MΦ-CL . Therefore these two methods are in agreement . The cause of actin depolymerisation in LD infection is unknown . It is tempting to speculate that intracellular LD by some unknown mechanism inhibits actin filament formation . Now the question comes how liposomal cholesterol treatment favors formation of actin filament . Recently we showed that liposomal cholesterol killed intracellular parasites [46] . Thus it may be possible that liposomal cholesterol treatment kills intracellular LD which may favor restoration of actin filament formation . The disruption of actin cytoskeleton may help to release the anchorage between cytoskeleton and membrane protein , making the movement of PLCδ1 free of cytoskeletal hindrance in LD infection . It may be recalled that a number of Salmonella strains carry spv virulence locus encoding the SpvB protein , an ADP-ribosyl transferase , which acts during intracellular infection to depolymerize the actin cytoskeleton [47] . There is also a report that Toxoplasma gondii infection changes the actin cytoskeleton of the dendritic cells due to secretion of parasite rhoptry [48] . In conclusion , this is the first report to our knowledge on changes in the cytoskeleton protein actin , coupled with increased lateral mobility of membrane protein in LD infected cells , which may have strong implication in altered plasma membrane architecture and defective signal transduction in infected macrophage .
|
The protozoan parasites , Leishmania donovani , replicate within the macrophages of the mammalian hosts . During its intracellular lifecycle , the parasite induces a wide variety of defects in the membrane homeostasis . Membrane bound receptor molecules are important for interacting with external stimuli . Our study very clearly showed that there is an increase in the mobility of membrane protein coupled with decrease in F-actin in infected cells , which may be corrected by liposomal cholesterol treatment . This observation indicates that intracellular parasite may alter the membrane biology of infected cells which may dampen overall cellular function .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"infectious",
"diseases",
"biology",
"and",
"life",
"sciences",
"parasitology",
"medicine",
"and",
"health",
"sciences"
] |
2014
|
Leishmania donovani Infection Enhances Lateral Mobility of Macrophage Membrane Protein Which Is Reversed by Liposomal Cholesterol
|
Taste is the primary sensory system for detecting food quality and palatability . Drosophila detects five distinct taste modalities that include sweet , bitter , salt , water , and the taste of carbonation . Of these , sweet-sensing neurons appear to have utility for the detection of nutritionally rich food while bitter-sensing neurons signal toxicity and confer repulsion . Growing evidence in mammals suggests that taste for fatty acids ( FAs ) signals the presence of dietary lipids and promotes feeding . While flies appear to be attracted to fatty acids , the neural basis for fatty acid detection and attraction are unclear . Here , we demonstrate that a range of FAs are detected by the fly gustatory system and elicit a robust feeding response . Flies lacking olfactory organs respond robustly to FAs , confirming that FA attraction is mediated through the gustatory system . Furthermore , flies detect FAs independent of pH , suggesting the molecular basis for FA taste is not due to acidity . We show that low and medium concentrations of FAs serve as an appetitive signal and they are detected exclusively through the same subset of neurons that sense appetitive sweet substances , including most sugars . In mammals , taste perception of sweet and bitter substances is dependent on phospholipase C ( PLC ) signaling in specialized taste buds . We find that flies mutant for norpA , a Drosophila ortholog of PLC , fail to respond to FAs . Intriguingly , norpA mutants respond normally to other tastants , including sucrose and yeast . The defect of norpA mutants can be rescued by selectively restoring norpA expression in sweet-sensing neurons , corroborating that FAs signal through sweet-sensing neurons , and suggesting PLC signaling in the gustatory system is specifically involved in FA taste . Taken together , these findings reveal that PLC function in Drosophila sweet-sensing neurons is a conserved molecular signaling pathway that confers attraction to fatty acids .
The gustatory system is critical for interpreting the nutritional value and potential toxicity of food compounds prior to ingestion . Nutritionally relevant food components are detected through specialized taste receptors expressed in sensory neurons that are broadly tuned to specific taste modalities [1]–[3] . Flies are also capable of detecting the caloric content of sugars through satiation feedback from internal sensors [4]–[7] . Taste represents an analytic sense , and unlike olfaction or vision , distinct taste modalities are sensed and processed independently from each other . The Drosophila gustatory system is divided into two main functional pathways that either detect appetitive sugars or aversive bitter substances [1] , [8] , [9] . Of the five basic taste qualities described in humans - sweet , sour , salty , umami , and bitter , fruit flies have been shown to detect tastants encompassed by only three of these taste modalities - sugars , bitter and salt [7] , [10] , [11] . Foods containing sugars , dietary lipids , and amino acids represent significant energy sources , and their presence tends to be attractive and promote consumption . In mammals , dietary lipids signal through mechanosensory and olfactory neurons , as well as postingestive feedback [12]–[15] . Dietary lipids are comprised of both triacylglycerides and fatty acids ( FAs ) , and growing evidence suggests that it is the free fatty acids that are detected by the gustatory system [16]–[23] . Fat represents a potent food source that yields more than twice the amount of energy as sugars per unit of mass . An understanding of how dietary FAs are sensed will provide critical insight into feeding choice and gustatory processing . While much is known about the detection and processing of sweet and bitter tastants in Drosophila , the neural basis for fat taste is unclear . Drosophila detect short-chain saturated FAs in free walking paradigms and they prefer low , while avoiding high FA concentrations [24] . Here we show that detection of a variety of FAs by the fly gustatory system induces a robust feeding response . These FAs serve as a dietary supplement with a potency that is comparable to sugars . FAs are perceived as appetitive at low and medium concentrations , and aversive at high concentrations . FA perception is independent of the olfactory system and acidity and instead requires the same gustatory sensory neurons that detect sugars . In mammals , phospholipase C ( PLC ) signaling is a critical second messenger required for taste . Our results demonstrate that PLC is uniquely required to sense FAs in Drosophila , revealing a conserved gustatory pathway that is independent from that required for sugar signaling .
To determine whether dietary fatty acids are sufficient for survival , flies were fed a diet composed exclusively of FAs ( Hexanoic acid – HxA , Octanoic acid – OcA , or Linoleic acid – LiA ) . HxA and OcA are short-chain saturated FAs that are naturally found in animal and plant products , including goat milk and coconut oil , and that are in the diet of some Drosophila species [24] . LiA is a long-chain unsaturated FA that is essential for human diet . The feeding preference and survival on FA diet was measured in a capillary feeding assay ( CAFE ) . Approximately 30–60 wild-type flies were starved for 48 hours prior to being placed in a vial with two capillary tubes: one containing 1% solution of various FA , and the other water . The number of surviving flies was measured over the course of 24 hours . Flies fed on FAs had a higher survival rate after 24 hours than control flies feeding on water alone ( P<0 . 01 for all concentrations , ANOVA; Fig . 1 ) . A dose-response curve revealed that low concentrations of HxA prolong survival in previously starved Drosophila . Flies were offered 1% , 0 . 4% or 0 . 1% solution of HxA and the numbers of surviving flies were measured over the course of 24 hours ( Fig . S1 ) . Flies fed 1% HxA showed no lethality throughout the length of the experiment . Flies fed 0 . 4% and 0 . 1% HxA showed a progressively decreasing survival rate that negatively correlated with concentration . For all concentrations of HxA tested , flies survived longer than control flies feeding on water alone ( P<0 . 01 for all concentrations ) . Taken together , these findings suggest that dietary FAs are metabolizable and partially sufficient for survival . When provided a choice in the CAFE assay ( Fig . 2A ) between FAs and water , flies strongly preferred FAs ( HxA , OcA , and LiA ) at concentrations of 0 . 1% or greater ( P<0 . 001 for all groups; Fig . 2B ) . Additionally , we found that flies display robust preference for oleic ( mono-unsaturated , omega-9 ) , decanoic and myristic acids ( both saturated FAs ) at concentrations of 0 . 4% in the CAFE assay ( data not shown ) . Dietary sugars are detected through gustatory receptors on the tarsi and proboscis as well as through internal metabolic sensors [5] , [6] , [25] , [26] . To investigate whether flies detect fatty acids through the peripheral gustatory system or through internal nutrient sensors , we measured the reflexive feeding response in Proboscis Extension Reflex ( PER ) assay ( Fig . 2A ) . Briefly , a small volume of either OcA or HxA was applied to the fly tarsi , and PER was measured as previously described [27] , [28] . When measuring PER , the tastant does not touch the proboscis , and therefore , cannot be ingested . Presentation of HxA or OcA dilutions ranging from 1% - 0 . 01% resulted in robust PER that was significantly greater than the response to water ( P<0 . 001 for all groups , except P<0 . 01 for 0 . 01% HxA ) , suggesting that peripheral gustatory receptors are sufficient for detection of FAs ( Fig . 2C ) . In Poxn mutant flies , external chemosensory sensillae are converted to mechanosensory sensillae [29] . These mutants can detect nutrients through internal sugar receptors , but do not display gustatory responses to tastants [6] . The PER response to 0 . 4% HxA ( as well as to sugars and yeast ) was abolished in Poxn mutant flies , further indicating that FAs are detected through peripheral sensory receptors ( Fig . 2D ) . The dietary sugars sucrose and fructose are strong gustatory attractants [30] . We sought to determine if flies can distinguish between FAs and sugars by testing whether flies exhibit concentration-dependent FA/sugar preference . To determine the sucrose response threshold , flies were provided a choice between water and sucrose in concentrations ranging from 0 . 1 to 5 mM in the CAFE assay and total ingestion was measured . Flies displayed strong preference for sucrose at 0 . 5 mM and higher ( P<0 . 001 for all groups ) ( Fig . 2E ) . When offered a choice between 0 . 4% HxA , or OcA , and a range of sucrose concentrations , flies preferred FAs over sucrose at concentrations less than 1 mM ( P<0 . 001 for sucrose 0 . 1 mM and 1 mM ) , while sucrose was preferred at concentrations greater than 2 mM ( P<0 . 001 for all groups for sucrose at 2 mM and 5 mM; Fig . 2E ) . These results reveal that flies display a concentration-dependent preference for FAs over sucrose . To determine whether concentration-dependent FA/sugar choice is specific to sucrose , we measured feeding preference comparing 0 . 4% HxA to a range of fructose concentrations . We found that flies similarly preferred HxA over fructose concentrations less than 1 mM ( P<0 . 001 for fructose 0 . 5 mM and 1 mM ) and fructose at concentrations greater than 2 mM ( P<0 . 001 for fructose 2 mM and 5 mM ) ( Fig . S2 ) . Taken together , these findings reveal that at certain concentrations , flies prefer FAs over sugars as a food source . Flies detect food through olfactory neuron dendrites that localize to the antennae and maxillary palps , and through gustatory neurons in the proboscis and legs [31]–[33] . These chemosensory organs are located relatively close to each other and are used for multimodal sensory processing of food cues [34] . To determine whether detection of FAs occurs independently from the primary olfactory system , we surgically removed antennae and maxillary palps , generating anosmic flies that lack olfactory organs [34] , [35] ( Fig . 3A ) . No significant differences were observed in the PER response to HxA , sugars ( fructose and sucrose ) or yeast extract between intact flies and flies lacking olfactory organs ( AntMxp-; P>0 . 05 , t-test for each pair; Fig . 3B ) . Preference for low concentration of HxA ( 0 . 01% ) and avoidance of a high concentration of HxA ( 5% ) in the CAFE assay did not differ between anosmic and intact flies ( 0 . 01% HxA P>0 . 568 , 5% HxA P>0 . 406 ) , suggesting olfaction is not required for HxA feeding preference or avoidance ( Fig . 3C ) . Taken together , these findings indicate that FA attraction is independent of the primary olfactory system . Fruit flies can sense acids and we sought to determine whether gustatory FA detection is dependent on acidity [36] . We tested preference for 1% HxA and OcA , as well as 0 . 1% acetic acid and 0 . 01% HCl ( pH∼3 . 5–3 . 6 for all ) in the CAFE assay . We also measured preference for the base NaOH ( pH∼9 . 5 ) to determine if high pH affects preference . Flies strongly preferred both HxA and OcA to water ( P<0 . 001 for both groups ) . Flies also preferred acetic acid ( P<0 . 008 ) but the preference was significantly lower than preference to FAs ( Fig . 3D , P<0 . 01 for both FAs ) . No significant preferences were observed with HCl ( P>0 . 094 ) or NaOH ( P>0 . 660; Fig . 3D ) suggesting that flies are generally not attracted to acidic or basic substances ( Fig . 3D ) . We tested the same concentrations of HxA and OcA against HCl in the CAFE assay . Despite matching pH , flies robustly preferred HxA and OcA over HCl ( P<0 . 001 to both FAs ) , suggesting that FA taste is mediated through chemical structure rather than low pH ( Fig . 3E ) . To directly measure whether acidity is required for FA taste , we adjusted the pH of 0 . 1% HxA to neutral ( pH∼7–7 . 2 ) by adding PBS buffer ( pH 7 . 4 ) . Flies strongly preferred pH-neutral HxA to PBS , confirming that FA taste is independent of acidity ( P<0 . 001; Fig . 3E ) . Flies sense sugars through gustatory receptor neurons that express gustatory receptor 64f ( Gr64f ) and can be labeled with Gr64f-GAL4 ( Fig . 4A ) , and aversive tastants through bitter-sensing neurons labeled by Gr66a-GAL4 [8] , [37] , [38] . These complementary populations of gustatory neurons can be selectively silenced through expression of the inward rectifying K+ channel Kir2 . 1 [39] . We expressed Kir2 . 1 under control of Gr64f-GAL4 to determine whether sweet-sensing neurons also detect FAs . To avoid potential developmental defects caused by silencing neurons throughout development , Kir2 . 1 expression was limited to adulthood with GAL80ts [40] , [41] . Briefly , adult-specific Kir2 . 1 expression was induced in sweet-sensing neurons by incubating 3 day-old flies at the non-permissive temperature of 30°C for 72 hours prior to testing . Flies were then starved for 48 hours at 22°C . Flies expressing Kir2 . 1 in sweet-sensing neurons and control flies were all tested at 22°C to prevent confounds of testing temperature on feeding behavior ( Fig . 4B ) . Silencing sugar-sensing neurons ( Gr64f-GAL4>UAS-Kir2 . 1 , GAL80ts ) abolished PER response to fructose and sucrose while control flies displayed robust PER ( P<0 . 001 compared to all controls , Fig . 4C ) . Strikingly , silencing Gr64f neurons also abolished PER response to all tested concentrations of HxA ( P<0 . 001 compared to all controls ) , indicating that Gr64f-expressing neurons are also required for HxA sensing ( Fig . 4C ) . Control flies of the same genotype ( Gr64f-GAL4>UAS-Kir2 . 1 , GAL80ts ) maintained at 22°C do not express Kir2 . 1 , and PER response to sugars or HxA was normal ( p>0 . 05 compared to other control groups , p<0 . 001 to the same genotype at 30°C ) . These findings indicate that FAs are sensed by , and confer feeding through , the same population of gustatory neurons that detect sugars . In vertebrates , the tastes of sweet , bitter , and amino acids are dependent upon phospholipase C ( PLC ) signaling [42]–[44] . We measured PER in response to FAs in flies mutant for no receptor potential A ( norpA ) , a fly ortholog of mammalian PLC . The mutant norpAP24 is a null allele and has previously been reported to have deficits in visual performance [45] . norpAP24 flies displayed dramatically reduced PER in response to HxA and OcA compared to wild-type controls ( P<0 . 001 for both groups ) , suggesting that norpA is required for FA taste ( Fig . 4D ) . However , PER response to fructose , sucrose , and yeast were comparable in norpAP24 and control flies ( P>0 . 05 for all groups ) , suggesting that norpA activity is required for sensing FAs specifically ( Fig . 4D ) . To localize the neurons where norpA is required for FA taste , we selectively restored norpA function to the sweet-sensing neurons . Flies with norpA expression limited to the Gr64f-expressing neurons showed greater PER response to HxA than norpAP24 mutants ( P<0 . 001 for both HxA concentrations ) and were statistically indistinguishable from control flies ( Gr64f-GAL4 , 0 . 4% HxA P = 0 . 808 and 1% HxA P = 0 . 082 ) . These findings suggest that norpA functions in sweet-sensing neurons to detect FAs ( Fig . 4E ) . No rescue was observed in flies with norpA expression limited to the rhodopsin-1 expressing neurons , where norpA is required for proper function of a visual system or in bitter-sensing Gr66a-expressing neurons ( Fig . S3 ) , confirming that the rescue of norpA in sweet- sensing neurons is not due to leakiness of the rescue transgene . To confirm rescue results , norpA was selectively targeted in sweet-sensing neurons through expression of two-independent RNAi lines . Transgenic flies with Gr64f-GAL4 and norpA-IR1 or norpA-IR2 displayed significantly reduced PER to HxA compared to control flies harboring Gr64f-GAL4 or UAS-RNAi transgenes alone ( Fig . S4; P<0 . 01 ) , confirming that norpA is required in sweet-sensing neurons for FA taste . Both sucrose and fructose response of flies with RNAi-norpA expressed under control of Gr64f-GAL4 was comparable to controls confirming that norpA expression in sweet-sensing neurons is selectively required for FA sensing . The receptors TRPM5 and TRPA1 signal through the PLC gustatory pathway in mammals and are proposed to be a polyunsaturated FA sensor in Drosophila and mammals [46] , [47] . In Drosophila , TRPA1 is also expressed in bitter-tasting neurons and confers avoidance of electrophiles [48] , [49] . However , TRPA1 mutant flies ( dTrpA1ins ) display a wild-type response to FAs suggesting TRPA1 is dispensable for FA taste in Drosophila ( Fig . S3 ) [50] . We conclude that FA taste in flies requires norpA/PLC function in sweet-sensing neurons , indicating that fly FA taste utilizes a pathway conserved in mammals .
Our findings demonstrate that FAs are sensed by the primary gustatory system and promote feeding . Flies displayed preference for 6 different FAs tested including hexanoic acid , octanoic acid , decanoic acid , myristic acid , linoleic acid and oleic acid . These represent diverse classes of FAs including short chain and long chain saturated FAs ( C6:0 to C14:0 ) as well as mono- and poly-unsaturated FAs ( C18:1 , C18:2 ) . These FAs were selected because of known preference by other species of Drosophila ( short-chain SFAs ) , preference by D . melanogaster larvae and adults ( long-chain saturated and unsaturated FAs ) or involvement in mosquito's olfactory preference cues ( long-chain SFAs ) [24] , [52] , [53] . Flies displayed robust responses to all FAs indicating that they are capable of sensing , and displaying preference for diverse FAs . Flies with surgically ablated olfactory organs retain robust appetitive response to FAs in CAFE and PER assays , showing that the preference for FAs is fully independent of the olfactory system ( Fig . 3B and C ) . High concentrations of FAs are aversive to flies and inhibit feeding through the gustatory and olfactory systems ( Fig . 3C ) . At high concentrations , the majority of short-chain FAs emits a pungent smell that is repulsive to Drosophila melanogaster . Species with unique host-plant preference including D . sechellia that feed on ripe Morinda citrifolia fruit show preference even to high concentration of short chain FAs [54] , suggesting that FA preference/avoidance choice is species-specific and dependent on diet . However , our findings reveal that low concentrations of short chain FAs induce a robust feeding response in D . melanogaster , which we demonstrated using two independent gustatory assays ( Fig . 2 ) . We employed the PER assay where only tarsal neurons are stimulated to distinguish between gustatory stimulation and ingestion of FAs . Robust appetitive response to FAs in the tarsal PER assay indicates that post-ingestive feedback is dispensable for detection and preference to FAs ( Fig . 2C ) . Preference for sugars based on nutritional information is sufficient even in the absence of gustatory cues [4]–[6] suggesting that peripheral sensory neurons and internal satiation sensors function independently . It remains to be determined whether flies are capable of sensing FAs through internal metabolic sensors . Future studies examining long-term food choice in norpA and Poxn mutant flies lacking FA taste may address this question . Fatty acids are hydrophobic chemicals and their texture differs from water or hydrophilic sugar solutions . Flies with genetically silenced gustatory neurons ( Gr64f-GAL4>UAS-Kir2 . 1 , GAL80ts ) do not respond to FAs or sugars ( Fig . 4C ) . Genetic silencing of sugar-sensing neurons does not impair mechanoreceptor function , indicating that the mechanical properties of FAs do not contribute to the FA-induced feeding response . Acid sensing in Drosophila regulates egg-laying , food-choice , and avoidance behavior [24] , [36] , [52] , [55] . However , flies robustly respond to HxA buffered to pH∼7 indicating that the appetitive response to FAs is independent of acidity . In mammals , FAs are detected through mechanosensory , gustatory and olfactory sensory systems [21] , [56] , [57] . Due to this multi-modal detection , establishing perception of dietary lipids and FAs as a distinct taste modality has been challenging [58] , [59] . Previous studies have revealed that D . melanogaster can detect FAs , but did not discriminate between feedback from internal satiation sensors , gustatory , or olfactory signals [24] , [52] . Our findings demonstrate that FAs are sensed specifically through the gustatory system , independent of acidic properties , mechanical , olfactory , or metabolic feedback . Therefore , in addition to sweet , bitter , salt , water and carbonation , FAs represent a novel taste modality in Drosophila [60]–[63] . FAs sensing requires the same neurons that detect sugars and induce feeding behavior . Genetic silencing of Gr64f neurons abolished PER response to all concentrations of HxA and all tested sugars ( Fig . 4C ) . The appetitive response elicited by FA-driven activation of sugar-sensing neurons indicates that these neurons harbor receptors for multiple taste modalities . In addition to sugars and FAs , the same neurons are activated by glycerol , an appetitive and nutritionally relevant alcohol that is detected through the specific receptor Gr64e [64] . The co-expression of multiple appetitive gustatory receptors allows Drosophila to categorize food sources in the absence of distinct neurons for each appetitive taste modality . Taken together , these findings support the labeled lines model for gustatory processing , where one subset of sensory neurons confers attractive behavior and the complementary subset confers repulsive behavior [9] , [60] . While it is clear that FAs are sensed in gustatory neurons , our findings do not rule out the presence of internal FA receptors . GRs mediating sugar-response are expressed in peripheral sensory neurons , but also in abdominal neurons where they are involved in detection of sugars in hemolymph and in metabolic regulation [25] , [65] , [66] . Flies can detect and respond to FA-based diet by perception of FAs through their peripheral sensory neurons , but it remains to be determined whether the internal neurons can also perceive FAs and regulate metabolically-relevant processes directly . Mutation of the PLC ortholog norpA abolishes the appetitive response to FAs , without affecting response to other appetitive taste stimuli including sugars and yeast . Expressing the wild-type allele of norpA selectively in sweet-sensing neurons under the control of Gr64f-GAL4 revealed that these neurons are necessary for detection of FAs , and the PLC signaling pathway is selectively required for FAs response . These findings indicate that shared neurons regulate FA and sugar taste , while distinct transduction pathways are involved in processing of each sensation . The Drosophila gene norpA is an essential component of the transduction pathways in visual and olfactory system [67] and has previously been implicated in TRPA1-dependent taste through function in bitter-sensing neurons [48] . The Drosophila genome encodes for two norpA isoforms [68] . It is possible that these isoforms have distinct functions that allow for independent regulation of vision and taste . In mice , PLC is selectively expressed in taste cells , and PLC knockout mice do not respond to sweet , amino acid , and bitter tastants [42] , [69] . The specific requirement for PLC signaling in FA taste in fly suggests a conserved gustatory transduction pathway that is more similar to mammalian taste than to other taste modalities in Drosophila . PLC-signaling is coupled to diacylgylcerol ( DAG ) that activates Drosophila Transient Receptor Potential ( TRP ) and TRP-like ( TRPL ) channels [70] , raising the possibility that TRP channels function as FA receptors . dTRPA1 functions in the Drosophila brain as a temperature sensor [50] and in the proboscis where it mediates avoidance response in bitter-sensing neurons [48] , [49] , [71] . In mammals , TRPA1 expresses in taste cells [72] and also functions as a receptor for polyunsaturated fatty acid [47]; however , we find that TRPA1 mutant flies have normal appetitive response to FAs ( Fig . S3 ) . In mammals , CD36 , a lipid binding protein , is expressed in gustatory oral tissue and appears to be selectively involved in FA taste . CD36 knock-out animals show no preference for FAs but retain their preference for sugars [20] , [73] . CD36 is conserved in flies but it is expressed only in olfactory neurons and function in olfactory detection of pheromones that are FA-derived [74] . Future work determining the FA receptors that activate PLC signaling will be central to understanding FA taste in Drosophila . While our findings reveal the importance of PLC signaling in Drosophila , we did not identify the receptor ( s ) required for sensing FAs . A number of GRs have unknown ligands and are co-expressed with Gr5a/Gr64f including Gr61a and Gr61b-d , raising the possibility that these are ligands for FAs [3] . Targeting these receptors selectively in Gr64f-expressing GRNs and testing flies for FA response in the CAFE or PER assays may be useful for identifying the FA receptor . A bioinformatic approach has also been used to search for gustatory receptors in Drosophila . Microarray analysis for genes differentially expressed between Poxn mutants that lack all chemosensory sensillae and wild-type flies , led to the identification of pickpocket28 , a Drosophila water receptor [63] . We localize FA taste to sweet-sensing neurons and therefore it is feasible to apply cell-sorting techniques followed by expression analysis [75] to reveal candidate receptors signaling FA taste . Previous work demonstrated that Drosophila use a relatively simple system of categorizing tastes within a given modality , discriminating distinct sugars based on intensity but not quality [30] . Because FAs are sensed by the same neurons that detect sugars it is possible that flies can only distinguish between FAs and sugars based on concentration-dependent intensity . Alternatively , FAs could be discriminated based on distinct temporal signaling resulting from the different transduction pathway . A parallel system is utilized by bitter-sensing neurons , where certain bitter substances signal through G-protein coupled receptors ( GPCRs ) , and electrophilic tastants signal though TRPA1 channels [49] . Future studies examining FA-conditioned memories may provide insight into gustatory processing in Drosophila and advance our understanding of gustatory conditioning . Testing FAs , sugars and glycerol in conditioning discrimination assay [5] , [28] , [30] may reveal whether different chemical groups are perceived differently based on their chemical structures and underlying transduction pathways .
Drosophila stocks were maintained on standard cornmeal/agar/molasses medium at 25°C , 70% humidity , in a LD incubator with 12∶12 light/dark cycle . Experiments were performed with wild-type Canton-S flies ( From M . Heisenberg , Wuerzburg University ) and the following transgenic lines were used: Gr64f-GAL4 ( From J . Carlson , Yale University; [76] , Kir2 . 1-GAL4;GAL80ts ( From H . Tanimoto , MPI , Munich; [40] ) , w;norpAP24 , UAS-norpA ( From C . Schnaitmann , MPI , Munich ) , w;norpAP24 [45] , w-;;dTrpA1ins [50] . The RNAi lines used to target norpA were part of the Transgenic RNAi Project collection from JFRC/HHMI . Bloomington stock #31113 is referred to as norpA-IR#1 and stock #31197 is referred to as norpA-IR#2 [77] . All chemicals used for behavioral assays were purchased from Sigma Aldrich including fructose , sucrose , hexanoic acid , octanoic acid , linoleic acid , acetic acid , oleic acid , decanoic acid , myristic acid , HCl and NaOH . Yeast extract ( Bio-Rad , NitroBacter ) . FAs were first diluted in 80% ethanol in ratio 1∶10 , then further diluted in water . Control solutions were also mixed with ethanol to achieve the same final concentration of ethanol . HxA was diluted in PBS buffer to increase pH to 7 . 2 . It was then tested against PBS of pH 7 . 4 . pH was measured by SevenEasy pH Meter , Mettler Toledo , Columbus , OH . Statistical analyses were performed using InStat software ( GraphPad Software 5 . 0 Inc . ) . For PER experiments , most tested groups violated the assumption of the normal distribution . Therefore , all the data were analyzed with non-parametric statistics . All experiments include data from >20 flies . Each fly was sampled three times with the same stimulus . The response was binary ( PER yes/no ) , and these three responses were pooled for values ranging from 0 to 3 . Kruskal-Wallis test ( nonparametric ANOVA ) was performed on the raw data from single flies and Dunn's Multiple Comparisons test was used to compare different groups . For capillary feeding assay , 30–60 flies were used per tube and 4 to 20 tubes per group were tested . Wilcoxon signed rank test ( non-parametric ) with two-tailed P value was used to test significance on single groups . In figures , graph bars are mean values and error bars are standard error of the mean .
|
The gustatory system is largely responsible for interpreting the nutritional value and potential toxicity of food compounds prior to ingestion . The receptors and neural circuits mediating the detection of sweet and bitter compounds have been identified in fruit fly , but neural mechanisms underlying detection of other taste modalities remain unclear . Here , we demonstrate through multiple lines of inquiry that fatty acids represent an appetitive cue that is sensed through the primary gustatory system . We find that fatty acids are detected by the same neurons that are also sensitive to sugars . Remarkably , the phospholipase C pathway , which mediates gustatory perception in mammals , is required in Drosophila for the taste of fatty acids but not sugars or bitter substances . Our findings reveal , for the first time , that fruit flies are capable of fatty acid taste , and identify a conserved molecular signaling pathway that is required for fatty acid feeding attraction .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2013
|
Drosophila Fatty Acid Taste Signals through the PLC Pathway in Sugar-Sensing Neurons
|
Open reading frame ( ORF ) 45 of Kaposi's sarcoma-associated herpesvirus ( KSHV ) is a tegument protein . A genetic analysis with a null mutant suggested a possible role for this protein in the events leading to viral egress . In this study , ORF45 was found to interact with KIF3A , a kinesin-2 motor protein that transports cargoes along microtubules to cell periphery in a yeast two-hybrid screen . The association was confirmed by both co-immunoprecipitation and immunoflorescence approaches in primary effusion lymphoma cells following virus reactivation . ORF45 principally mediated the docking of entire viral capsid-tegument complexes onto the cargo-binding domain of KIF3A . Microtubules served as the major highways for transportation of these complexes as evidenced by drastically reduced viral titers upon treatment of cells with a microtubule depolymerizer , nocodazole . Confocal microscopic images further revealed close association of viral particles with microtubules . Inhibition of KIF3A–ORF45 interaction either by the use of a headless dominant negative ( DN ) mutant of KIF3A or through shRNA-mediated silencing of endogenous KIF3A expression noticeably decreased KSHV egress reflecting as appreciable reductions in the release of extracellular virions . Both these approaches , however , failed to impact HSV-1 egress , demonstrating the specificity of KIF3A in KSHV transportation . This study thus reports on transportation of KSHV viral complexes on microtubules by KIF3A , a kinesin motor thus far not implicated in virus transportation . All these findings shed light on the understudied but significant events in the KSHV life cycle , delineating a crucial role of a KSHV tegument protein in cellular transport of viral particles .
Kaposi's sarcoma–associated herpesvirus ( KSHV ) , also known as the human herpesvirus 8 ( HHV-8 ) , is a human DNA tumor virus [1] . KSHV is etiologically associated with the endothelial neoplasm Kaposi's sarcoma ( KS ) and with certain lymphoproliferative disorders like primary effusion lymphoma ( PEL ) and multicentric Castleman's disease ( MCD ) [2] , [3] . KSHV infection of cells by default establishes latency . During this phase , there is expression of only a limited number of the viral ( latent ) genes essential for maintenance of the viral genome with no production of infectious virions [4] . Disruption of latency results in the reactivation of the virus into the lytic phase , with expression of the entire viral gene panel and production of infectious viral particles [5] , [6] . These events thus ensure the propagation and transmission of viruses to uninfected cells serving to maintain the infection [7] . Lytic phase is also essential in sustaining the population of latently infected cells that otherwise would be quickly lost by segregation of latent viral episomes as spindle cells divide [8] . The KSHV lytic phase and constant primary infection of fresh cells is thus crucial for both the viral tumorigenicity and the disease pathogenesis . With the exception of viral DNA replication and viral gene expression , other events that follow viral reactivation including virus assembly , transportation and egress however have been much less studied in KSHV . To gain more knowledge about these events it thus becomes essential to identify the different virion proteins and recognize their functional roles . For this purpose , we purified extracellular KSHV virions from tetra deconyl phorbol acetate ( TPA ) -induced BCBL-1 cells ( that are latently infected with KSHV ) by double gradient ultracentrifugation and identified the component virion proteins by a mass spectrometric analysis . By this approach a total of 24 different virion-associated proteins were identified by a mass spectrometric analysis including five capsid proteins , eight envelope glycoproteins and eleven tegument or putative tegument proteins [9] . Among the different virion proteins , the tegument proteins in KSHV as well as in other related gamma herpesviruses have not been well studied . Current knowledge of these proteins comes mainly from studies on alpha and beta-herpesviruses wherein they contribute to three essential functions in the viral life cycle . First , some serve a regulatory function , modulating the host cellular environment during the immediate-early phase of infection like the virion host shutoff protein ( UL41 ) of HSV-1 that degrades host mRNA and shuts down the host translation program [10] . Second , some play vital roles in transportation of capsids to the nucleus along microtubules following virus entry into the host cell [11]–[13] . Third , tegument proteins also participate in the complex chain of events involving herpesviral assembly and egress including transportation of viral complexes [14] , [15] . Based on these evidence , the need arises to investigate the functional roles if any of even the less studied KSHV tegument proteins . One among these proteins in KSHV is ORF45 , also recognized as an immediate-early ( IE ) protein [16] , [17] . A vital role of ORF45 in events leading to viral egress following virus reactivation emanated from two observations . First , we had generated an ORF45-null recombinant KSHV using the bacterial artificial chromosome ( BAC ) system by inserting a premature stop codon into the ORF45 coding sequence [18] . Upon reconstitution into a 293T cell system followed by induction , this mutant produced a much lower yield of progeny virions compared to the wild-type virus though viral gene expression and DNA replication remained unaffected [18] . This suggested that ORF45 could possibly have a role to play in the stages subsequent to viral DNA replication and viral protein synthesis , mostly involving transportation of viral complexes toward viral egress . The second supporting observation came from a yeast two hybrid ( Y2H ) screening employed to identify ORF45 interacting cellular partners that revealed ORF45 to interact with cDNAs of KIF3A ( in this study ) . KIF3A is a subunit of kinesin-2 , a microtubule ( MT ) plus-end-directed motor protein . Kinesin-2 comprises two motor subunits , KIF3A and either KIF3B or KIF3C , and a non-motor subunit , KAP3 ( kinesin superfamily-associated protein-3 ) [19]–[21] . Kinesin-2 molecules are expressed ubiquitously and transport cargoes along microtubules ( MTs ) from the nucleus to the cell periphery [19] , [20] , [22] . KIF3A comprises of an N-terminal head ( motor ) domain that attaches to and migrates along MTs and a C-terminal tail that presumably functions as the cargo-binding domain [20] , [23] . In this study we thus explored the role and significance of ORF45–KIF3A interaction in the possible transportation of KSHV viral complexes along microtubules toward egress following reactivation from latency . We found that ORF45 through its interaction with the cargo-binding domain of KIF3A docked the entire viral capsid-tegument complexes onto KIF3A which were subsequently transported along microtubules for viral maturation and egress . This to the best of our knowledge is the first report attributing a definitive role of KIF3A in virus transportation .
Earlier studies with an ORF45-null mutant virus revealed that disruption of ORF45 yielded 10-fold lowered titers of progeny virions as compared to the wild-type virus though viral gene expression and DNA replication remained unaffected [18] . This finding suggested a role of ORF45 in virion assembly or egress . To further explore the function of ORF45 in these processes , we attempted to identify cellular proteins interacting with it . By an Y2H screening we identified interferon regulatory factor-7 ( IRF-7 ) as an ORF45-associated protein with ORF45 antagonizing interferon- related anti-viral responses [24] . Upon revisiting the Y2H results , another cellular protein was also identified as an ORF45-binding protein . This was KIF3A , a subunit of kinesin-2 and a microtubule ( MT ) plus-end-directed motor protein . KIF3A , a 702 amino acid protein contains three domains , head ( motor ) , stalk and tail . Two prey plasmids isolated from the screening contained cDNAs for the C-terminal fragments of KIF3A with both fragments encompassing amino acids 409 to 702 . Interaction between ORF45 and KIF3A in cells was further confirmed by co-immunoprecipitation ( co-IP ) experiments . ORF45 and KIF3A expression vectors were used to co-transfect 293T cells . Forty-eight hours post-transfection , cells were lysed and immunoprecipitated with an anti-ORF45 antibody . By a Western blot , KIF3A was found to be immunoprecipitated with ORF45 , suggesting that the two proteins are physically associated in a complex in cells ( data not shown ) . The interaction of these two proteins was also examined in BCBL-1 cells latently infected with KSHV . Cell lysates prepared from TPA induced BCBL-1 cells and immunoprecipitated with a mouse monoclonal anti-ORF45 antibody revealed KIF3A in the immunoprecipitate detected with a rabbit anti-KIF3A antibody ( Figure 1A , lane 3 ) . As a control , KIF3A was not immunoprecipitated with mouse IgG ( Figure 1A , lane 2 ) . A reverse co-IP performed on cell lysates prepared from TPA-induced BCBL-1 cells with a rabbit-polyclonal anti-KIF3A antibody revealed immunoprecipitation of ORF45 with KIF3A ( data not shown ) . This thus proved the interaction between these two proteins in the reverse direction , too . Furthermore , the localization of ORF45 and KIF3A in BCBL-1 cells was also examined by immunoflorescence assay ( IFA ) . BCBL-1 cells induced with TPA for 48 hours were fixed and reacted with mouse monoclonal anti-ORF45 and rabbit-polyclonal anti-KIF3A antibodies . Both proteins were exclusively localized in the cytoplasm ( Figure 1B ) . The staining of ORF45 and KIF3A completely overlapped , again suggesting that these proteins are associated with each other in the cytoplasm of the cells . To further characterize the interaction between the two proteins , we attempted to map the domains of ORF45 interacting with KIF3A by Y2H and co-IP approaches . For the Y2H assay , full-length and a series of ORF45 truncation/deletion mutants in pACT2 vector ( prey ) were co-transformed with full-length KIF3A in pAS2-1 vector ( bait ) into yeast cells . Yeast transformants positive for prey-bait interaction were selected on plates lacking leucine , tryptophan and histidine but containing 3-AT of an appropriate concentration and assayed for β-galactosidase activity . The amino ( amino acids 1–115 ) and carboxy ( amino acids 332–407 ) termini of ORF45 interacted with KIF3A while the central domain ( amino acids 115–332 ) did not interact with KIF3A ( Figure 2A , left panel ) . For the co-IP assay , a series of ORF45 truncation/deletion mutants were cloned in pCMV–Tag2 vector to express Flag-tagged fragments . The truncation/deletion mutants were introduced into 293T cells by transfection . KIF3A was not included in the transfection as it is ubiquitously expressed in all cells . Upon confirmation of optimal and stable expression of the full-length and the truncation mutants of ORF45 by a Western analysis , cell lysates were immunoprecipitated with anti-Flag M2 affinity gel , followed by a Western blot with an anti-KIF3A antibody . KIF3A was immunoprecipitated with all the truncation mutants of ORF45 from the C-terminus ( mutants 1–115 , 1–237 , 1–332 , 1–383 ) ( Figure 2A , right panel , lane 2 , lanes 5–7 ) though with reduced affinities compared to the full-length ORF45 ( Figure 2A , right panel , lane 1 ) . A similar finding was also seen with respect to truncations of ORF45 from the N-terminus ( mutants 238–407 , 19–407 , 77–407 , 90–407 , 115–407 ) ( Figure 2A , right panel , lane 4 , lanes 8–11 ) . The central domain ( amino acids 115–332 ) however did not interact with KIF3A ( Figure 2A , right panel , lane 3 ) . These findings thus indicated only the amino ( 1–115 ) and the carboxy ( 332–407 ) terminals of ORF45 to interact with KIF3A , consistent with the Y2H findings . We next mapped domains of KIF3A interacting with OF45 . KIF3A , a 702 amino acid protein consists of an amino terminal head ( motor ) domain ( aa 1–355 ) , a central stalk ( aa 356–587 ) and a cargo binding domain at the carboxy terminal tail ( aa 588–702 ) ( Figure 2B ) . The motor domain has both the ATP and the microtubule binding sites . The tail ( cargo binding ) domain associates with cargoes transporting them along MTs . Mapping was performed by an Y2H assay , wherein the prey ( KIF3A truncation/deletion segments in pACT2 vector ) and the bait ( full-length ORF45 in pAS2-1 vector ) were co-transformed into yeast cells . Yeast transformants positive for prey-bait interaction were selected as above and assayed for β-galactosidase activity . The KIF3A truncation segments spanning amino acids 1–400 ( constituting only the motor domain ) , 400–600 ( constituting only the central stalk ) and 1–600 ( containing only the motor domain and the stalk ) failed to interact with ORF45 . However , the segment spanning amino acids 600–702 ( constituting the cargo-binding domain ) interacted with ORF45 . Additionally , deletions of the motor domain ( segment 400–702 with Δ 1–399 ) or the stalk ( 1–702 with Δ 400–500 or Δ 500–600 ) did not block the affinity of KIF3A to ORF45 . But absence of the cargo-binding domain ( 1–600 with Δ 601–702 ) abolished the interaction of KIF3A with ORF45 ( Figure 2B ) . These results suggested that ORF45 specifically interacted with the cargo-binding domain of KIF3A . ORF45 specifically binds to the KIF3A subunit of Kinesin-2 and also has been shown to be tightly associated with KSHV tegumented capsids through interaction with many other tegument proteins including ORF64 and ORF63 and capsid proteins ORF 62 [25] . Thus we hypothesized that ORF45 may mediate association of assembling viral particles to the motor molecules and play a role in their transportation along microtubules during virion maturation . To validate the model , we asked if KIF3A associates with only ORF45 or with a whole capsid-tegument complex . We addressed this question by immunoprecipitating cell extracts obtained from TPA induced 293T cells harboring wild-type KSHV ( BAC36 ) with an anti-KIF3A antibody . Upon analyzing the immunoprecipitates , we found that in addition to ORF45 , other tegument proteins ( ORFs 33 and 64 ) and capsid proteins ( ORFs 62 and 65 ) were co-precipitated with KIF3A ( Figure 3A , lane 4 ) . The viral envelope glycoproteins ( gB and gH ) however failed to co precipitate with KIF3A ( Figure 3A , lane 4 ) . The whole cell extracts ( WCE ) showed the expression of all the analysed proteins ( Figure 3A , lane 1 ) . A similar experiment performed on induced BCBL-1 cells also revealed identical findings ( data not shown ) . Thus these observations suggested that KIF3A is associated and thereby involved in the transportation of only the entire viral capsid-tegument complexes and not the envelope glycoproteins which are transported separately from tegument-capsid particles in the cytoplasm . We next investigated the specific role of ORF45 in associating viral particles to KIF3A . There were two possibilities ( 1 ) ORF45 could just be one among the many KSHV virion proteins interacting with KIF3A; ( 2 ) ORF45 probably mediates the association of viral complexes with KIF3A . Hence to delineate the role of ORF45 in this process , we performed a parallel immunoprecipitation assay on cell lysates collected from induced 293T cells carrying an ORF45-null mutant ( BAC-stop45 ) . In the absence of expression of ORF45 both the tegument and the capsid proteins failed to coprecipitate with KIF3A ( Figure 3A , lane 6 ) , providing a strong evidence that the viral capsid-tegument complexes docked onto KIF3A through ORF45 . The lack of co-precipitation of tegument-capsid proteins with KIF3A in the ORF45-null mutant also could have been due to a defect in tegument-capsid complex assembly as a result of lack of functional ORF45 protein . To investigate this posibility , we also performed a co-IP with anti-ORF65 antibody with cell lysates from both BAC36 and BAC-stop45-carrying cells , followed by Western analyses with specific antibodies to some virion proteins . Intact viral capsid-tegument complexes as evidenced by presence of ORFs 65 , 62 , 33 , and 64 were detectable in the immunoprecipitates obtained from both wild-type ( Figure 3B , lane 4 ) and the ORF45-null mutant ( Figure 3B , lane 6 ) viruses . Taken together , these results proved that the absence of functional ORF45 does not disrupt viral tegument-capsid complex assembly processes , but impairs the loading of the viral complexes on kinesin-2 molecules . Studies with other herpesviruses like HSV and pseudorabies virus ( PrV ) have shown the requirements for an intact MT network to ensure transport of viral complexes toward cell periphery for viral egress [11] , [26]–[28] . Additionally , from the association of the KSHV capsid-tegument complexes with KIF3A , a motor protein moving along microtubules , it only seemed logical to investigate the role of MTs in the transportation of the viral complexes following reactivation . To address this issue , BCBL-1 cells were treated with increasing doses ( 0 . 5 , 5 . 0 , 10 . 0 µM ) of nocodazole , a microtubule depolymerizing drug [29] , followed by induction of viral lytic cycle with TPA . Four days post-induction , virions were pelleted and viral copy numbers estimated by a real time PCR . With increasing concentration of nocadazole in a nontoxic range , there was an appreciable decrease in the extracellular virion titers down to a level of 3 . 5×105 copies/ml amounting to a 8 fold reduction compared to the levels obtained with nocadazole untreated and induced BCBL-1 cells ( Figure 4 ) . Hence damage to the MT architecture through depolymerization by nocadazole had contributed to inefficient transport of the viral complexes reflecting as reduced viral titers . To ensure that the decrease in viral titers seen was not due to the toxic effects of the drug permanently damaging the cellular or the virion architecture , we also performed a nocodazole removal ( washout ) experiment in parallel . Herein cells treated with 5 µM of the drug were washed twice with PBS and once with serum-free medium to ensure removal of the drug followed by induction with TPA . Such drug removal studies have shown to result in rapid repolymerization of MTs within a short time span [30]–[32] . In our study following the drug removal , a reversal of its effect was evidenced by build up of virion titer levels comparable to that of the untreated but induced cells ( Figure 4 ) . Thus drug removal had resulted in MT repolymerization , resuming efficient transport of the viral complexes . To ensure that the decrease in viral titers was not due to nocodazole treatment irreversibly damaging the virion architecture or preventing viral tegument-capsid complex formation , nocodazole-treated and TPA-induced BCBL-1 cells were immunoprecipitated with anti-ORF65 antibody . In addition to ORF65 , other capsid ( ORF62 ) and tegument ( ORFs 45 , 64 , and 33 ) proteins were detected in the immunoprecipitates by Western blot with specific antibodies , indicating the presence of intact tegument-capsid complexes in nocodazole-treated cells ( data not shown ) . This proved that nocodazole did not damage the virion architecture . To further investigate if KSHV particles are associated with MTs , a double-labeled IFA was performed . TPA induced BCBL-1 cells were fixed and permeablized . MTs and viral capsids were detected with mouse monoclonal anti-tubulin and rabbit-polyclonal anti-ORF65 antibodies respectively with anti-mouse IgG-Alexa Flour 488 ( green ) and anti-rabbit–IgG-Texas Red ( red ) secondary antibodies . Upon confocal microscopic examination , a series of optical sections of the image were collected at 0 . 32 µm intervals from the bottom to the top of the image . At each interval , the two channels were recorded sequentially and/or simultaneously and images acquired . The individually acquired MT ( green channel ) and the viral capsid ( red channel ) staining in a cell are shown in Figure 5A and 5D , and 5B and 5E , respectively , representing two different optical sections at which images were acquired . The merged image shows viral capsids as red structures localized along the MTs ( Figure 5C and 5F ) . A reconstructed 3-D image of the cell constructed revealed close association of viral particles with MTs ( Video S1 ) . IFA was also performed on BCBL-1 cells following nocodazole treatment . Results are shown in Figure 5G–5I and Video S2 . First of all , the absence of microtubule network was shown with anti-tubulin antibody staining in the cells treated with nocodazole . Second , we expected to see majority of the viral particles localized only around the perinuclear region . But as shown in Video S2 , absence of staining with anti-tubulin prevented us from getting a definitive conclusion . In addition , the cell morphology is another reason for the difficulty to locate the viral particles within the cytoplasm as the nuclear membrane could have been only separated by a few micrometers from the plasma membrane . Thus the viral particles seen could actually have been localized to the perinulear zone , but due to the above phenomenon it might have been difficult to demonstrate it . However , images of the cell taken at different ‘Z’ depths do in fact show viral particles to be more concentrated toward the center of the cell ( i . e . , the nucleus ) . Further IFA performed on cells following nocodazole removal gave a picture similar to that of the drug-untreated control with viral particles associated with microtubules ( data not shown ) . To investigate the significance of this interaction in the KSHV life cycle , we employed a dominant negative ( DN ) mutant approach . Such approaches have been often used for functional studies of kinesins , including kinesin-2 [33]–[35] . One among the DN mutants , is a headless mutant , with the head ( motor ) domain deleted . This mutant is neither able to hydrolyze ATP nor bind to microtubules hence unable to migrate . The KIF3A fragment ( amino acids 409–702 ) , isolated from our Y2H screening as interacting with ORF45 , could theoretically work as a DN mutant of KIF3A . This fragment is unable to migrate along microtubules as it lacks the N-terminal motor domain ( spanning amino acids 1–355 , Figure 6A ) but could still compete with wild type KIF3A in binding to ORF45 ( and thereby viral complexes ) due to the intact C-terminal cargo-binding domain . First , to test if the KIF3A-DN mutant binds to viral particles , BCBL-1 cells were electroporated with the KIF3A-DN mutant plasmid ( cloned into a pCMV-flag tagged vector ) along with RTA-expression vector ( pCR3 . 1-ORF50 plasmid ) . Forty-eight hours post-transfection , cell lysates obtained were immunoprecipitated with anti-flag M2 Affinity gel . In addition to ORF45 , other virion tegument ( ORFs 64 , 33 ) and capsid proteins ( ORFs 62 , 65 ) were detected by a Western analysis along with the flag-tagged KIF3A-DN protein . However , the DN protein was not able to bind to microtubules as seen by the absence of alpha-tubulin in the immunoprecipitate ( Figure 6B ) . This suggested that though the DN mutant is bound to viral particles it still cannot transport them along microtubules due to the absence of the N-terminal motor domain . We then tested if this DN mutant is able to block KSHV particle assembly and release . The KIF3A DN mutant at increasing concentrations ( 0 . 5 , 1 . 0 , 2 . 0 and 10 . 0 µg ) was introduced into BCBL-1 cells by electroporation along with pCR3 . 1-ORF50 plasmid for induction into the lytic phase . Four days post-transfection , extracellular viral particles were collected and quantified by a real-time PCR [18] . With increasing inputs of the KIF3A DN mutant , there was a noticeable decrease in the extracellular virion titers down to a level of 3 . 7×105 copies/ml , amounting to a 7-fold reduction compared to levels obtained with DN-untreated but -induced BCBL-1 cells ( Figure 6C ) . KIF3A DN mutant thus effectively inhibited the KIF3A-ORF45 interaction indicating that the ORF45-mediated interaction of viral particle to kinesin-2 is crucial for the transport of particles toward viral egress . To rule out the possibility that the decrease in virion titers seen with the KIF3A DN mutant could possibly have been due to its generalized cytotoxic effects on the cells or other detrimental effects on the viral components , a similar experiment was also performed in 293T cells infected with KSHV or HSV-1 . We hypothesized that HSV-1 could serve as an effective negative control based on earlier studies which have documented only the requirements of kinesin-1/kinesin heavy chain ( KHC ) protein and not that of KIF3A in the transportation of HSV viral complexes toward egress [36]–[38] . Using an RNA interference approach , we also showed that KHC indeed has a significant role in HSV-1 transportation toward egress and that KIF3A plays no discernable role in HSV-1 intracellular transportation ( Figure S1 ) . These data clearly illustrated that HSV-1 could be employed as an effective negative control for the KIF3A DN mutant studies . 293T cells transfected with increasing concentrations of the KIF3A DN mutant plasmid ( 0 . 5 , 1 . 0 , 2 . 0 and 10 . 0 µg ) were infected with either KSHV ( at 50 genomes per cell ) or with HSV-1 ( 5 Pfu/cell ) . With KSHV , following infection , cells were induced with TPA and virion titers estimated . With increasing inputs of the DN mutant , virion titers dropped to about 1 . 1×105 copies/ml amounting to a 9-fold decrease compared to levels obtained with DN untreated but induced 293T cells ( Figure 6D ) , similar to that seen with BCBL-1 cells . With HSV-1 , at three different time points ( 18 , 24 and 36 hours post-infection ) , extracellular virions were collected and titers measured by a real-time PCR . There was no noticeable decrease in titers even with increasing concentrations of the KIF3A DN mutant with levels similar to that of the DN untreated control ( Figure 6E ) . Introduction of the KIF3A DN mutant had no detrimental effect on the cells or HSV-1 suggesting that the decrease in viral titers seen with KSHV in both BCBL-1 and 293T cells was specifically attributable to the competitive effect of the KIF3A DN on wild-type KIF3A in binding to ORF45 . This clearly points out to the crucial role of KIF3A-ORF45 interaction in pathways leading to KSHV egress . Though the KIF3A DN mutant reduced the KSHV viral egress considerably , we did not notice a complete abrogation of viral egress . There could be two possibilities for this scenario: ( i ) though greatly reduced by the DN mutant , some full-length KIF3A could still interact with ORF45 mediating viral transportation; ( ii ) other kinesin motors may also contribute to transportation of viral complexes . To address the first possibility , we attempted to knockdown KIF3A expression in cells through a short-hairpin RNA ( shRNA ) -based approach and examine its effect on KSHV transportation and hence egress . A Mission shRNA gene set against human KIF3A was purchased from Sigma-Aldrich . The Mission shRNA system is a lentiviral vector based RNA interference library against annotated human genes , which generates siRNAs in cells and mediates gene specific RNA interference for extended periods of time . The KIF3A shRNA set consists of four individual shRNA lentiviral vectors against different target sites of KIF3A mRNA sequence ( clone #1 was directed against the 3′ UTR and the clone #s 2–4 targeting the coding sequence ) . After introduction into BCBL-1 cells by lentiviral transduction , all the four KIF3A shRNA clones were found to effectively down-regulate KIF3A expression to almost undetectable levels in comparison to the control ( Figure 7A ) . BCBL-1 cells stably expressing the KIF3A and control shRNAs , respectively , were induced with TPA . Four days post-induction extracellular virions were collected and titers estimated by a real-time PCR . All of the four KIF3A shRNA clones were effective in drastically reducing the extracellular virion titers down to a level of about 1×105 copies/ml amounting to a nearly 24 - 25 fold reduction compared to the controls ( BCBL-1 cells transduced with control shRNA/TPA induced and untransduced BCBL-1 cells/TPA induced ) ( Figure 7B ) . The reductions were more pronounced than that with the DN approach . To ensure that the decrease in viral titers was not due to any deleterious effects of the introduced lentiviruses or the shRNA sequences on KSHV replication , intracellular KSHV genomic DNA were analyzed at forty-eight hours post-induction by a real-time PCR . The viral genome copies were normalized to GAPDH . Levels of viral replication were not significantly altered in the lentivirus transduced cells compared to non-transduced/TPA induced cells ( Figure 7C ) . Hence the decrease in KSHV viral titers represents most likely a direct effect of KIF3A knockdown . Furthermore , this set of KIF3A shRNAs did not affect HSV-1 egress in 293T cells while the HSV-1 production was dramatically inhibited by three shRNAs that target KHC/kinesin-1 ( Figure S1 ) . Overall , the findings from the KIF3A knockdown studies clubbed with the conclusions from the DN mutant studies definitively point to a very pivotal role of KIF3A in transportation of KSHV tegumented capsids , establishing the specificity of KIF3A in KSHV transport . As to the second possibility that other kinesins may also contribute to KSHV intracellular transport , we also included a set of shRNAs against KHC ( kinesin-1 ) in the studies to look into the role of KHC in KSHV transport . KHC is the typified and the most widely studied among the kinesin group of proteins and it has been definitively implicated in transportation of related herpesviruses like HSV-1 [36]–[38] and many other viruses including vaccinia [39] and West-Nile virus structural proteins [40] . Three of four KHC shRNA clones effectively down-regulated KHC expression to almost undetectable levels as compared to the control ( Figure 7A and Figure S1A ) . None of the three KHC shRNA clones exhibited any drastic effects on KSHV egress to that seen with KIF3A but decreasing the viral egress very slightly ( just about 1-fold ) ( Figure 7B ) . This finding thus does not exclude the utility of KHC in KSHV transport and suggests possibly a very minor role for KSHV egress . Although it is difficult to conclude that kinesin-2 is the only motor molecule necessary and sufficient for KSHV egress as human kinesin family consists of at least 14 different members , our data indicate that KIF3A remains the primary mediator .
An ORF45-null mutant virus was profoundly defective in the release of extracellular mature infectious virions though with no obvious defects in overall gene expression and lytic DNA replication , suggesting a role of ORF45 in virion assembly and egress [18] . Compelling evidences presented in this report substantiate this possible role by revealing the requirement of ORF45 in the kinesin-2-mediated transport of assembled viral particles along microtubules after nuclear egress . The evidences are ( i ) ORF45 specifically interacts with the cargo-binding domain of KIF3A and is colocalized with the cellular motor protein in the cytoplasm; ( ii ) Entire viral tegument-capsid complex was associated with KIF3A with ORF45 mediating the association; ( iii ) Intact microbutules were required for transport of viral complexes toward egress as demonstrated by nocadazole treatment and IFA; ( iv ) Disruption of KIF3A-ORF45 interaction with a headless DN mutant of KIF3A or through knockdown of endogenous KIF3A expression by siRNAs noticeably decreased KSHV egress reflected as appreciable reduction in the release of extracellular virions; ( v ) Knockdown of KHC expression had no appreciable effects on the viral egress demonstrating the specificity of KIF3A in KSHV transportation . Based on these evidences a model for the role of ORF45 in virion maturation and egress has been presented ( Figure 8 ) . In this model , after nuclear egress , KSHV capsids acquire tegument proteins including ORF64 , ORF63 and ORF45 in the cytoplasm . ORF45 on the capsid-tegument particles recruits KIF3A and mediates loading of the viral particles onto kinesin-2 . Subsequently , viral particles are transported along microtubules in the cytoplasm from the perinuclear region to the cell periphery or trans-golgi network ( TGN ) membrane where envelopment and egress occur ( Figure 8 ) . Transportation and acquirement of envelope glycoproteins is independent to that of the KIF3A transportation of viral tegument-capsid complexes as evidenced in our study wherein KIF3A failed to associate with glycoproteins ( gB and gH ) . Similar findings have also been observed in HSV-1 where the anterograde transport of unenveloped capsids and the glycoproteins in axons involved dissociated pathways [27] , [36]–[38] , [41] . Similar to ORF45 , an identical role of tegument proteins in transportation of viral complexes has been shown in related herpesviruses like HSV-1 and PrV . Specific tegument proteins of HSV-1 like US11 and UL56 interact with related kinesins like the kinesin-1 motors playing vital roles in the anterograde transport of viral complexes along microtubules in the axons [36]–[38] , [42] . Outer tegument proteins pUL47 , pUL48 and pUL49 of PrV mediate its anterograde transport [11] . The major tegument protein pUL36 in both HSV and PrV contributes to the MT-dependent transport of capsids during egress [26] , [28] , [43] . Homologues of KSHV ORF45 though found in other members of gamma herpesviruses , are not present in alpha- and beta-herpesviruses . An earlier study from our laboratory revealed a role of ORF45 in antagonizing interferon antiviral defenses . KSHV ORF45 interacted with IRF-7 efficiently suppressing virus-mediated interferon ( IFN ) gene expression [24] . But the ORF45 counterparts of Epstein-Barr virus ( EBV ) and Rhesus monkey Rhadinovirus ( RRV ) did not interact with IRF-7 ( our unpublished data ) , suggesting that interaction with IRF-7 and inhibition of its activation is probably a function unique to KSHV ORF45 . In contrast , KIF3A in addition to KSHV ORF45 also interacted with its analogues in EBV and RRV by both Y2H and co-IP ( data not shown ) . The second function of ORF45 in virus transport through interaction with KIF3A could thus be uniformly conserved across all gamma herpesviruses . An amino acid sequence alignment of ORF45 sequences from related gamma-herpesviruses performed with the ClustalW software ( NCBI website ) revealed a significant conservation of amino acids in the N-terminal and the C-terminal domains of the protein across the spectrum ( data not shown ) . This observation further substantiated our ORF45 mapping data in KSHV wherein we detected only these highly conserved domains ( amino acids 1–115 and 332–407; amino and carboxy terminals respectively ) of ORF45 to interact with KIF3A , while the sparsely conserved middle domain ( amino acids 115–332 ) failed to associate with KIF3A . The C-terminal region ( spanning amino acids 600–701 ) of KIF3A was found to interact with ORF45 . This region has been predicted to be the possible cargo-binding domain of the protein [20] , [23] . Similar interactions of viral proteins with the cargo-binding domain of other kinesins have also been reported in additional viruses including West-Nile virus [40] , HIV-1 [44] and other herpesviruses like HSV-1 . Interestingly in HSV-1 , US11 tegument protein interacted with the heptad repeat cargo-binding domain of Kinesin-1 [36] contributing to the anterograde transport of nucleocapsids along axons . Interactions of the cargo-binding domain of KIF3A with viral proteins have not been reported thus far . However other cargoes , like the PAR-3 protein ( a protein complex functioning in various cell polarization events ) [35] and GADD34 ( possibly involved in transcription and DNA recombination ) [23] specifically interacted with the cargo-binding domain of KIF3A . To assess if the KIF3A-ORF45 interaction is crucial for the virus in transportation toward viral egress , we further probed their interaction dynamics by two different approaches . The first approach involved the use of a DN mutant of KIF3A that lacks the motor domain but retains the cargo-binding domain . This mutant was seen to compete with full-length KIF3A in binding to ORF45 hence reducing the transport of viral complexes toward viral egress reflecting as a noticeable fall in virion release . Two valid observations arose from this approach: ( i ) KIF3A–ORF45 interaction is of definitive significance in the transportation of KSHV viral capsid-tegument complexes; ( ii ) it demonstrates a potential value of the DN mutant in a peptide therapy to treat KSHV-associated diseases which could be further investigated . With the DN mutant approach , we however did not notice a total abrogation of viral egress . This finding was expected , as the DN mutant can only competitively inhibit the full length KIF3A-ORF45 interaction and not totally abolish the interaction . We hypothesized that a much clearer picture on the contribution of KIF3A in transportation of viral complexes could only arise by approaches that tend to totally inhibit its interaction with ORF45 . This was made possible by attempting to knockdown the endogenous KIF3A expression in BCBL-1 cells . A lentivirus based approach was employed to generate cells harboring the KIF3A shRNA sequences integrated into the host cell genome . This allowed for the stable and long-term gene down regulation , overcoming the difficulties associated with only a transient gene knockdown following direct transfection of small interfering RNA ( siRNA ) . Stable knockdown of KIF3A expression manifested as a highly significant reduction in viral egress , more pronounced than that encountered with the DN approach . Human kinesins constitute a large family of motor proteins with at least 14 members . Our data through gene knockdown studies suggest that kinesin-2 is a major vehicle for KSHV intracellular transport while kinesin-1 ( KHC ) is crucially responsible for HSV-1 transport . Although the likelihood of other kinesins contributing to KSHV and HSV-1 transport in minor roles cannot be ruled out , the fact that KSHV and HSV-1 egress was clearly not compensated by other kinesins during the shRNA knockdown of KIF3A and KHC respectively exemplifies the specificity of kinesin-2 in KSHV transport and KHC in HSV-1 movement along microtubules . Though the observations seen with the shRNA mediated knockdown of KIF3A/KHC on KSHV egress , seem to be exciting , these however have to be interpreted cautiously . Down-regulation of the kinesins could possibly have other off-target effects on the viral or cellular components . This could be especially true given the fact that kinesins are involved in cellular processes . Possible deleterious effects on the viral processes associated with the knockdown were ruled out from the observation that the levels of viral DNA replication in cells expressing the KIF3A/KHC shRNAs were similar to that seen in cells expressing the negative control shRNA . More importantly , among the off-target effects on the cells likely to impact on KSHV egress , would be any effects on the microtubule structure . Earlier studies have shown that KIF3A knockdown in mammalian cells do not affect the microtubule architecture [45] . Thus with levels of viral DNA replication and the microtubule architecture remaining unaffected , we could safely conclude that the initial steps in virus assembly remain unaffected by the knockdown . KIF3A down regulation has been shown to produce slightly aberrant Golgi complex morphology in mammalian cells [45] . This is not likely to be the reason for a decrease in KSHV egress , as HSV-1 microtubule-dependent transport was not significantly affected . The association of KIF3A with the viral capsid-tegument complexes was principally mediated by ORF45 . Indeed , ORF45 was found to be tightly associated with the pelleted capsid complex when purified virions treated with detergents ( Triton X-100 plus 0 . 5% DOC ) were subjected to high-speed centrifugation [9] , [17] . In addition , the interactions of ORF45 with a multitude of tegument proteins including ORF64 and ORF63 and capsid proteins like ORF62 have been revealed in a virion-wide protein interaction study from our laboratory [25] . Thus ORF45 being associated tightly with the capsid-tegument complex is a probable component of the inner tegument acquired early following exit of capsids from the nuclear membrane hence is placed strategically to efficiently position the viral capsid-tegument complexes onto KIF3A . Further mapping studies by Y2H approaches , revealed that ORF45 interacted with other proteins like ORF64 and ORF 63 only through its middle segment ( spanning amino acids 115–332 ) ( our unpublished data ) . This suggested that ORF45 could possibly have 2 distinct interacting domains; the amino and carboxy terminal ends that interact with KIF3A and the central domain interacting with the other virion proteins . This possibly ensures a more efficient binding of KIF3A to its binding sites on ORF45 . In related herpesviruses like HSV-1 , recruitment of KHC by tegument proteins contributes to transportation of viral complexes toward viral egress [36]–[38] , [42] . Contrastingly our study findings point out that KSHV recruits KIF3A ( through ORF45 ) for transportation of viral capsid-tegument complexes . This finding seems highly intriguing , the most plausible explanation for the differences in the recruited kinesins could reside in the highly divergent tail ( cargo-binding ) domain of the kinesins that are known to specify distinct set of cargoes for each of them . Given the fact that ORF45 , a protein with no homologs in alpha and beta- herpesviruses interacts with KIF3A , it is tempting to speculate that KIF3A might be uniquely adapted to mediate transportation of KSHV and other gamma-herpesviruses . Two observations in our study substantiated the finding that microtubules serve as the major cellular highway for KIF3A to transport the viral capsid-tegument complexes . Treatment of BCBL-1 cells with a MT depolymerizer nocodazole [29] caused an appreciable decrease in virion production with drug removal replenishing the viral titer levels . Further confocal microscopic images revealed the association of viral particles with MTs following reactivation . Similar requirements of an intact microtubular structure for transport of virus complexes toward egress have been shown for other herpesviruses like HSV and PrV [11] , [26]–[28] , [38] . Revelation of the process of KSHV particle transport on microtubules during virion assembly provides novel strategies for halting viral replication and treating KSHV-associated diseases . For example , microtubule structure in cells can be disrupted by reagents like nocodazole ( an MT destablizer ) or taxol ( MT stablizer ) as many of such reagents are shown to suppress replication of various viruses . Our studies may lead to discovery of new drugs for KS and reveal pharmacological mechanism of the drugs . In this study we thus report on the transportation of KSHV viral complexes along microtubules by KIF3A , a kinesin motor which to the best of our knowledge has thus far not been implicated in virus transport . This finding could thus lead researchers to investigate similar roles of KIF3A in transportation of other viruses . Furthermore , our study by delineating the role of ORF45 has helped unravel the initial pathways of KSHV transportation toward egress . However further assembly and transportation of mature enveloped virions still remains an enigma . This final transportation phase could probably involve recruitment of kinesin motors either by the glycoptotein cytoplasmic tails or by membrane associated tegument proteins as suggested for HSV-1 [38] . On these lines , a recent study has identified an interaction of HSV-2 membrane-associated tegument protein pUL56 with KIF1A [37] . This interaction has been speculated to contribute to the axonal transport of viral glycoprotein-containing vesicles . Another study has revealed an association of KIF5B with enveloped HSV-1 containing in abundance the amyloid precursor protein ( APP ) , a kinesin cellular receptor [42] . Though these studies provide interesting observations , further investigation of this phenomenon is warranted . The sequence of events involved in transportation of KSHV viral complexes to the nucleus following cell entry also needs investigation with specific focus on the virion proteins and the cellular motors involved . Information garnered from all these sources would definitely help in providing the entire picture of the KSHV lytic life cycle starting from viral entry and ending with viral egress . Lastly , the significance of the less studied KSHV tegument has been exemplified by the pivotal and multi functional roles of ORF45 . This should pave way for future studies examining functional roles of even the other tegument proteins in the viral life cycle .
BCBL-1 , a latent KSHV-infected primary effusion lymphoma cell line was maintained in RPMI 1640 . Human embryonic kidney ( HEK ) 293T cells were maintained in Dulbecco's modified Eagle medium ( DMEM ) . All cultures were supplemented with 10% heat-inactivated fetal bovine serum ( FBS ) and antibiotics . The prey plasmids used for Y2H were constructed by cloning either the full-length or the different truncation/deletion mutants of KSHV ORF45 and KIF3A ( Figure 2A ) in pACT2 vector ( between the BamHI & XhoI sites ) in frame with the vector's GAL4 Activation Domain ( AD ) . The deletion mutants of both ORF45 and KIF3A were generated with a PCR-based mutagenesis system , ExSite ( Stratagene ) using a pair of phosphorylated oligonucleotides in opposite directions with either pACT2-ORF45 or the pACT2-KIF3A as the template plasmids respectively . Bait plasmids were constructed by cloning either the full-length KIF3A or the ORF45 sequences in pAS2-1 vector in frame with GAL4 DNA-binding domain ( DBD ) . For co-immunoprecipitation ( co-IP ) assays , full-length and the different truncation mutants of KSHV ORF45 ( Figure 2A ) were PCR amplified and cloned into pCMV-Tag2 vector ( Stratagene ) between the BamHI & XhoI sites . The deletion mutants of ORF45 were constructed using the ExSite ( Stratagene ) system with a pair of phosphorylated oligonucleotides using pCMV-Tag2-ORF45 as the template . The ORF45 and KIF3A prey plasmids were tested for their ability to interact with KIF3A and ORF45 bait plasmids respectively by a yeast-two hybrid screen , performed with the Matchmaker system ( CLONTECH ) . The haploid yeast strain MATα MaV103 ( gift from Dr . Marc Vidal at the Massachusetts General Hospital ) was co-transformed with prey and bait plasmids using lithium acetate . Yeast transformants positive for prey-bait interaction were selected on plates lacking leucine , tryptophan and histidine but containing 3-Amino-1 , 2 , 4-triazole ( 3-AT ) and subsequently assayed for β-galactosidase activity using the standard colony-filter assay ( as per the Clontech Yeast Protocols handbook ) . 293T cells grown in 100 mm dishes to 70% confluency were transfected with 10 µg of expression plasmid by the calcium phosphate transfection method . BCBL-1 cells were transfected by electroporation . Plasmid DNAs were mixed with BCBL-1 cells in OPTI-MEM medium ( Gibco-BRL ) and electroporated ( 200 V , 960 µF ) with a Genepulser II ( Bio-Rad , Hercules , Calif . ) . Electroporated cells were then transferred to RPMI 1640 medium supplemented with 10% serum and maintained for the time as indicated . BCBL-1 cells were induced with TPA ( 20 ng/ml ) for 48 hours . Cell lysates prepared in ice-cold lysis buffer [25] were homogenized and clarified by high-speed centrifugation at 4°C and subjected to an immunoprecipitation with mouse-monoclonal anti-ORF45 [9] or rabbit-polyclonal anti-KIF3A ( Sigma ) antibodies for 2 hours or overnight at 4°C . Immunoprecipitated complexes were thoroughly washed with cold lysis buffer , resuspended in 100 µl of sodium dodecyl sulfate-polyacrylamide gel electrophoresis ( SDS-PAGE ) loading buffer , boiled for 10 mins and loaded onto SDS-PAGE gels ( Invitrogen ) . The primary antibodies used for Western blotting were mouse monoclonal anti-ORF45 and rabbit-polyclonal anti-KIF3A . Subsequent incubation with secondary antibodies and detection was done as earlier [25] . Similarly , transfected 293T cells were collected forty-eight hours post-transfection , lysed with ice-cold lysis buffer and homogenized as above . Immunoprecipitation was performed as above by incubating lysates with anti-flag M2 Affinity gel ( A2220 , Sigma ) . Immunoprecipitated protein complexes were washed and run on a SDS-PAGE . Primary antibodies used for Western blotting were the mouse monoclonal anti-Flag M2 antibody ( Sigma ) and the rabbit-polyclonal anti-KIF3A ( Sigma ) . Incubation with secondary antibodies and subsequent detection was performed as earlier [25] . Frozen glycerol stocks of the BAC36 ( BAC-cloned wild-type KSHV ) and BAC-stop45 ( ORF45-null mutant virus ) [18] were retrieved and BAC DNA was prepared with the large construct kit ( QIAGEN ) . The freshly prepared DNA was transfected into 293T cells , stable BAC-293T monolayers generated through hygromycin selection and induced with TPA . Cells lysates collected forty-eight hours post-induction were incubated with rabbit-polyclonal anti-KIF3A antibody ( Sigma ) for 2 hours or overnight at 4°C , followed by incubation with Protein G agarose beads ( Invitrogen ) for 2 hours at 4°C . A reaction with 2 µl of rabbit IgG ( Sigma ) was included as a negative control . By Western blot , in addition to KIF3A , the IP complexes were also analysed for the different virion protein using specific antibodies including mouse monoclonal anti-ORF45 , rabbit-polyclonal anti-ORF64 and mouse polyclonal anti-ORF33 [9] , mouse anti-ORF62 IgM ( gift from Dr . Z . H . Zhou at UCLA ) , rabbit-polyclonal anti-ORF65 , rabbit-polyclonal anti-gB ( gift from Dr . Johnan Kaleeba at the Uniformed Services University of the Health Sciences ) , mouse monoclonal anti-gH ( gifts from Dr . Gary Cohen at Penn ) . BCBL-1 cells were treated with increasing concentrations ( 0 . 5 , 5 . 0 , 10 . 0 µM ) of the microtubule depolymerizing drug nocodazole ( Sigma ) at 37°C for 1 hour and then induced with 20 ng/ml TPA for 4 days . For nocadazole washout experiments , cells were treated with 5 µM of the drug , washed twice with PBS , once with serum-free medium to ensure complete removal of the drug and then induced with TPA . Post-induction , virions were purified and pelleted from the medium supernatant as detailed earlier [9] and were resuspended in 1× phosphate-buffered saline ( PBS ) in 1/100 of the original volume . Concentrated viruses were first treated with Turbo DNase I ( Ambion ) at 37°C for 1 h to remove any contaminating DNA outside viral particles . Viral DNA was liberated by digestion with lysis buffer ( AL ) and proteinase K ( supplied with the DNeasy tissue kit , Qiagen ) and extracted with phenol-chloroform . Extracted DNA was precipitated with ice-cold ethanol , and the final DNA pellet was dissolved in TE buffer . Copy numbers of KSHV genomic DNA were estimated by real-time DNA PCR with a Roche LightCycler and the LightCycler FastStart DNA MasterPlus SYBR green kit with primers directed to LANA [46] . Viral DNA copy numbers were calculated from external standards of known concentrations of BAC36 DNA . A serial dilution of a known amount of BAC36 DNA was used to construct a standard curve . Copy numbers were normalized and were expressed as copy number per milliliter of supernatant . BCBL-1 cells were induced with TPA ( 20 ng/ml ) for 48 hours . Cells were washed with PBS , fixed with 2% paraformaldehyde in PBS , and spun onto Shandon double cytoslides at 1 , 000 rpm for 5 min . Cells then were permeabilized in 0 . 2% Triton X-100 in PBS for 20 min on ice and subjected to a double-labeled IFA with mouse monoclonal anti-ORF45 [9] and rabbit-polyclonal anti-KIF3A ( Sigma ) antibodies . Fluorescein isothiocyanate ( FITC ) -conjugated anti-mouse IgG and Texas Red-conjugated anti-rabbit IgG ( Vector Laboratories , Inc . ) were used as the respective secondary antibodies . Slides were examined with a Leica TCS SPII confocal laser scanning system . The 2 channels were recorded simultaneously and/or sequentially and controlled for possible breakthrough between the green and red channels . For visualization of the viral particles along MTs , induced , fixed and permeablized BCBL-1 cells were subjected to a double labeled IFA . MTs were stained with mouse monoclonal anti-tubulin antibody ( Sigma ) whilst the viral particles were stained with rabbit-polyclonal anti-ORF65 ( capsid antibody ) . Goat anti-mouse IgG-Alexa Flour 488 ( Molecular Probes , Invitrogen ) and goat anti-rabbit–IgG-Texas Red ( Vector Laboratories ) were used as the respective secondary antibodies . Double staining was then examined under a confocal microscope ( Leica TCS SPII confocal laser scanning system ) by the use of the 495- and 590-nm bands of laser lines from a water-cooled argon-krypton laser under an oil immersion objective . Channel recording and control of data acquisition was done as earlier . A series of optical sections of the image were collected at increasing intervals of 0 . 32 µm from the bottom to the top . This was employed for the reconstruction of a 3-D representation of the specimen on the graphics computer with the LCS 3D software ( Leica Microsystems ) . Digital images obtained were cropped and adjusted for contrast with Photoshop . The C-terminal region of KIF3A ( spanning amino acids 409–702 ) was PCR amplified and cloned into the pCMV 3-Tag- ( flag ) vector between BamHI and XhoI sites . This was designated the KIF3A DN mutant . Increasing concentrations ( 0 . 5 , 1 . 0 , 2 . 0 or 10 . 0 µg ) of this mutant plasmid along with 10 µg of pCR3 . 1-ORF50 [47] or empty pCR3 . 1 vector were mixed and electroporated into BCBL-1 cells as mentioned earlier and maintained for 4 days . ORF50 encodes for the replication transctivator ( RTA ) essential for induction of cells into the lytic phase [48] . 293T cells were also transfected with increasing concentrations ( 0 . 5 , 1 . 0 , 2 . 0 , 10 . 0 µg ) of the KIF3A DN plasmid by the calcium phosphate transfection method . Following transfection , cells were infected with concentrated KSHV ( at 50 genomes per cell ) plus Polybrene ( 4 µg/ml ) . Virus inocula was then removed , cells were washed twice and replaced with fresh medium containing FBS . Forty-eight hours following infection monolayers were induced with 20 ng/ml TPA ( for 4 days ) . Extracellular KSHV virions were purified and collected from either induced BCBL-1 or 293T cells . DNA extraction from intact virions and subsequent KSHV genomic DNA quantitation was performed as detailed earlier . KIF3A DN mutant transfected 293T cells were also infected with HSV-1 ( 5 Pfu/cell ) and incubated at 37°C ( for 1 hour ) . Cells were washed twice and replenished with fresh medium containing FBS . At different time points ( 18 , 24 and 36 hours ) following infection , the medium was clarified , extracellular virions were collected and DNA isolated . Copy numbers of HSV-1 genomic DNA was estimated by a real-time PCR with primers directed against the HSV-1 UL30 gene [49] . For external standards , HSV-1 genomic DNA was extracted from the HSV-1 KOS strain ( kindly provided by Dr . Gary Cohen at Penn ) and quantitated . A serial dilution of a known concentration of this DNA was used to construct a standard curve . Two Mission shRNA gene sets against human KIF3A and KIF5B/KHC respectively were purchased from Sigma-Aldrich . The KIF3A set consists of four individual shRNA lentiviral vectors in pLKO . 1-puro plasmids against different target sites of KIF3A ( with Clone IDs NM_007054 . 4-2737s1c1 , -2134s1c1 , -781s1c1 , -1945s1c1 – for convenience sake referred as KIF3A shRNA clone #s 1 , 2 , 3 , 4 ) . The KIF5B/KHC set contains five clones targeting different sites of KIF5B/KHC ( with Clone IDs NM_004521 . 1-3461s1c1 , -1376s1c1 , -1904s1c1 , -391s1c1 , -842s1c1 – referred as KHC shRNA clone #s 1 , 2 , 3 , 4 , 5 ) . A control vector ( a non-targeting shRNA that activates the RNAi pathway without targeting any known human gene , SHC002 ) was also purchased ( Sigma-Aldrich ) . Each of the shRNA vectors as well as the control vector was used to prepare lentiviral stocks by cotransfecting 293T cells with the shRNA vector and two packaging vectors ( pHR'8 . 2ΔR and pCMV-VSV-G ) at a ratio of 4∶3∶1 respectively . . Three days post-transfection , the culture media that contain shRNA retroviruses were harvested , centrifuged ( 500×g for 10 min at 4°C ) , and filtered through a 0 . 45 µm filter to ensure removal of any non adherent cells . BCBL-1 cells were transduced with the shRNA encoding lentivirus stocks in the presence of polybrene ( 8 µg/ml ) . Transduced cells were selected with puromycin ( 2 µg/ml ) for a week . Efficacies of these shRNAs in knockdown of the respective proteins were assayed by Western blot with specific antibodies . BCBL-1 cells stably expressing the KIF3A/KHC/control shRNAs were treated with TPA ( 20 ng/ml ) to induce KSHV lytic replication . Four days post-induction , extracellular KSHV virions were collected . DNA extraction from intact virions and viral titer estimation were performed as detailed earlier . Sub-confluent 293T monolayer cells were also transduced with the lentivirus stocks in the presence of polybrene and selected with puromycin as above . 293T cells stably expressing the KIF3A/KHC/control shRNAs were infected with HSV-1 at an MOI of 5 . Twenty-four hours post-infection , the supernatant medium was clarified , extracellular HSV-1 virions were pelleted and viral DNA isolated . Viral titers were estimated by a real-time PCR as earlier .
|
Kaposi's sarcoma–associated herpesvirus ( KSHV ) is a tumor virus associated with Kaposi's sarcoma ( KS ) and a spectrum of other lymphomas . These tumor cells are usually latently infected with this virus . The inactive virus in cells can get reactivated , whereupon there is viral DNA replication and viral protein synthesis . Newly synthesized proteins assemble in an orderly fashion to form viral complexes that need to be transported to the cell periphery for release and to further infect fresh cells to maintain the infection . Events that make up this important phase in the viral life cycle , however , have been much less studied . In this study , we show that a KSHV protein called the open reading frame ( ORF ) 45 anchors newly assembled viruses onto a cellular motor protein , namely KIF3A . These viruses are then transported by KIF3A along microtubules which act as major cellular highways ( tracks ) , allowing for efficient transportation of viral complexes toward the cell periphery . Inhibition of any of these steps resulted in a reduced transport of viral complexes reflecting as reduced viral levels . Thus , this study has helped to delineate crucial events involved in the transportation of newly assembled KSHV virions and provides for attractive viral and cellular targets that could be inhibited to reduce the virus burden .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"virology/virion",
"structure,",
"assembly,",
"and",
"egress",
"cell",
"biology",
"virology"
] |
2009
|
Kaposi's Sarcoma-Associated Herpesvirus ORF45 Interacts with Kinesin-2 Transporting Viral Capsid-Tegument Complexes along Microtubules
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.